Professional Documents
Culture Documents
Ghosh S. Mathematics and Computer Science Vol 2. 2023
Ghosh S. Mathematics and Computer Science Vol 2. 2023
Science Volume 2
Scrivener Publishing
100 Cummings Center, Suite 541J
Beverly, MA 01915-6106
Series Editors: Niranjanamurthy M, PhD, Juanying XIE, PhD, and Ramiz Aliguliyev, PhD
Scope: Data engineering is the aspect of data science that focuses on practical applications of data collection and
analysis. For all the work that data scientists do to answer questions using large sets of information, there
have to be mechanisms for collecting and validating that information. Data engineers are responsible for
finding trends in data sets and developing algorithms to help make raw data more useful to the enterprise.
It is important to have business goals in line when working with data, especially for companies that handle
large and complex datasets and databases. Data Engineering Contains DevOps, Data Science, and Machine
Learning Engineering. DevOps (development and operations) is an enterprise software development phrase
used to mean a type of agile relationship between development and IT operations. The goal of DevOps is
to change and improve the relationship by advocating better communication and collaboration between
these two business units. Data science is the study of data. It involves developing methods of recording,
storing, and analyzing data to effectively extract useful information. The goal of data science is to gain
insights and knowledge from any type of data — both structured and unstructured.
Machine learning engineers are sophisticated programmers who develop machines and systems that can
learn and apply knowledge without specific direction. Machine learning engineering is the process of
using software engineering principles, and analytical and data science knowledge, and combining both
of those in order to take an ML model that’s created and making it available for use by the product or
the consumers. “Advances in Data Engineering and Machine Learning Engineering” will reach a wide
audience including data scientists, engineers, industry, researchers and students working in the field of
Data Engineering and Machine Learning Engineering.
Publishers at Scrivener
Martin Scrivener (martin@scrivenerpublishing.com)
Phillip Carmical (pcarmical@scrivenerpublishing.com)
Mathematics and Computer
Science Volume 2
Edited by
Sharmistha Ghosh
M. Niranjanamurthy
Krishanu Deyasi
Biswadip Basu Mallik
and
Santanu Das
This edition first published 2023 by John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
and Scrivener Publishing LLC, 100 Cummings Center, Suite 541J, Beverly, MA 01915, USA
© 2023 Scrivener Publishing LLC
For more information about Scrivener publications please visit www.scrivenerpublishing.com.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or other-
wise, except as permitted by law. Advice on how to obtain permission to reuse material from this title
is available at http://www.wiley.com/go/permissions.
For details of our global editorial offices, customer services, and more information about Wiley prod-
ucts visit us at www.wiley.com.
ISBN 978-1-119-89632-6
Set in size of 11pt and Minion Pro by Manila Typesetting Company, Makati, Philippines
10 9 8 7 6 5 4 3 2 1
Contents
Preface xvii
1 A Comprehensive Review on Text Classification and
Text Mining Techniques Using Spam Dataset Detection 1
Tamannas Siddiqui and Abdullah Yahya Abdullah Amer
1.1 Introduction 2
1.2 Text Mining Techniques 3
1.2.1 Data Mining 3
1.2.2 Information Retrieval 4
1.2.3 Natural Language Processing (NLP) 5
1.2.4 Information Extraction 5
1.2.5 Text Summarization 6
1.2.6 Text Categorization 7
1.2.7 Clustering 7
1.2.8 Information Visualization 7
1.2.9 Question Answer 8
1.3 Dataset and Preprocessing Steps 9
1.3.1 Text Preprocess 9
1.4 Feature Extraction 9
1.4.1 Term Frequency – Inverse Document Frequency 10
1.4.2 Bag of Words (BoW) 10
1.5 Supervised Machine Learning Classification 11
1.6 Evaluation 11
1.7 Experimentation and Discussion Results for Spam
Detection Data 11
1.8 Text Mining Applications 13
1.9 Text Classification Support 13
1.9.1 Health 13
1.9.2 Business and Marketing 14
1.9.3 Law 14
v
vi Contents
1.10 Conclusions 14
References 15
2 Study of Lidar Signals of the Atmospheric Boundary Layer
Using Statistical Technique 19
Kamana Mishra and Bhavani Kumar Yellapragada
2.1 Introduction 19
2.2 Methodology 21
2.2.1 A Statistical Approach to Determine the CBLH 21
2.2.2 A Statistical Approach to Determine the Best
Fit Distribution to the Backscatter Signals of the
Lidar Dataset 22
2.3 Mathematical Background of Method 23
2.4 Example and Result 24
2.5 Conclusion and Future Scope 27
Acknowledgement 28
References 28
Annexure 30
3 Optimal Personalized Therapies in Colon Cancer Induced
Immune Response using a Fokker-Planck Framework 33
Souvik Roy and Suvra Pal
3.1 Introduction 33
3.2 The Control Framework Based on Fokker-Planck Equations 35
3.3 Theoretical Results 39
3.4 Numerical Schemes 41
3.5 Results 43
3.6 Conclusion 44
Acknowledgments 45
References 45
4 Detection and Classification of Leaf Blast Disease using
Decision Tree Algorithm in Rice Crop 49
Sarvesh Vishwakarma and Bhavna Chilwal
4.1 Introduction 49
4.2 Proposed Methodology 51
4.3 Result Analysis 52
4.4 Conclusion 55
4.5 Future Work 55
References 56
Contents vii
The mathematical sciences are part of nearly all aspects of everyday life.
The discipline has underpinned such beneficial modern capabilities as
internet searching, medical imaging, computer animation, weather predic-
tion, and all types of digital communications. Mathematics is an essential
component of computer science. Without it, you would find it challenging
to make sense of abstract language, algorithms, data structures, or differen-
tial equations, all of which are necessary to fully appreciate how computers
work. In a sense, computer science is just another field of mathematics.
It does incorporate various other fields of mathematics, but then focuses
those other fields on their use in computer science. Mathematics matters
for computer science because it teaches readers how to use abstract lan-
guage, work with algorithms, self-analyze their computational thinking,
and accurately model real-world solutions. Algebra is used in computer
programming to develop algorithms and software for working with math
functions. It is also involved in design programs for numerical programs.
Statistics is a field of math that deploys quantified models, representations,
and synopses to conclude from data sets.
This book focuses on mathematics, computer science, and where the
two intersect, including heir concepts and applications. It also represents
how to apply mathematical models in various areas with case studies. The
contents include 29 peer-reviewed papers, selected by the editorial team.
xvii
1
A Comprehensive Review on Text
Classification and Text Mining Techniques
Using Spam Dataset Detection
Tamannas Siddiqui and Abdullah Yahya Abdullah Amer*
Abstract
Text data mining techniques are an essential tool for dealing with raw text data
(future fortune). The Text data mining process of securing exceptional knowledge
and information from the unstructured text is a fundamental principle of Text data
mining to facilitate relevant insights by analyzing a huge volume of raw data in
association with Artificial Intelligence natural language processing NLP Machine
Learning algorithms. The salient features of text data mining are attracted by the
contemporary business applications to have their extraordinary benefits in global
area operations. In this, a brief review of text mining techniques, such as cluster-
ing, information extraction, text preprocessing, information retrieval, text clas-
sification, and text mining applications, that demonstrate the significance of text
mining, the predominant text mining techniques, and the predominant contem-
porary applications that are using text mining. This review includes various exist-
ing algorithms, text feature extractions, compression methods, and evaluation
techniques. Finally, we used a spam dataset for classification detection data and a
three classifier algorithm with TF-IDF feature extraction and through that model
achieved higher accuracy with Naïve Bayes. Illustrations of text classification as
an application in areas such as medicine, law, education, etc., are also presented.
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (1–18) © 2023 Scrivener Publishing LLC
1
2 Mathematics and Computer Science Volume 2
1.1 Introduction
Text data mining techniques are predominantly used for extracting relevant
and associated patterns based on specific words or sets of phrases. Text data
mining is associated with text clustering, text classification, and the prod-
uct of granular taxonomy, sentiment analysis, entity relation modeling, and
document summarization [1]. Prominent techniques in text mining tech-
niques include extraction, summarization, categorization, retrieval, and
clustering. These techniques are used to infer distinguished, quality knowl-
edge from text from previously unknown information and different written
resources obtained from books, emails, reviews, emails, and articles with
the help of information retrieval, linguistic analysis, pattern recognition,
information extraction, or information extraction tagging and annotation
[2]. Text preprocessing is the predominant functionality in text data min-
ing. Text preprocessing is essential to bring the text into a form that can be
predictable and analyzable for text data mining. Text preprocessing can be
done in different phases to formulate the text into predictable and analyz-
able forms. These are namely lowercasing, lemmatization, stemming, stop
word removal, and tokenization. These important text preprocess steps are
predominantly performed by machine learning algorithms for natural lan-
guage processing tasks. These preprocessing steps implement data cleaning
and transformation to eliminate outliers and make it standardized to cre-
ate a suitable model to incorporate the text data mining process [3]. Text
data mining techniques are predominantly used for records management,
distinct document searches, e-discovery, organizing a large set of a text data,
analysis and monitoring of understandable online text in internet commu-
nication and blogs, identification of large textual datasets associated with
patients during a clinical area, and clarification of knowledge for the readers
with more extraordinary search experience [4]. Text data mining techniques
are predominantly used in scientific literature mining, business, biomedical,
and security applications, computational sociology, and digital humanities
as shown in Figure 1.1 below.
Dimensionality
reduction
The paper reviews text data mining techniques, various steps involved
in text preprocessing, and multiple applications that implement text data
mining methods discussed in Table 1.1.
Text
collection
Preprocessing
Text data
Feature Selection
Unstructured Structured
Web Text Sequences
Sign of the Zodiac:
1. Aries
2. Taurus
The second sign of the Zodiac is 3. Gemini...
Taurus. Most Common Cause of
Strokes are the third most common cause of Death in America:
1. Heart Disease
death in America today. 2. Cancer
No study would be complete without 3. Stroke...
mentioning the largest rodent in the Largest rodent in the
world, the Capybara. world:
1. Capybara
2. Beaver
3. Patagonian Cavies
1.2.7 Clustering
In clustering, the main objective is to perform cluster analysis of content-
based text records. It uses ML and NLP to recognize and classify unstruc-
tured text data using extracted descriptors from a targeted document
available in the database. Word clustering is part and parceling of text
clustering for part sets of information in subsets of semantically related
words. It is aimed to perform duties ranging from word sense or basic dis-
ambiguation to knowledge retrieval [22]. Unsupervised text clustering is
widely used in NLP to group similar data. It obtains the distance between
the points. The predominant clustering methods are namely soft clustering
and hard clustering. Hard clustering performs the grouping task for every
object that belongs to exactly one cluster. Hard clustering allows an object
to be grouped with one or more clusters based on its meaning and nature
[23].
classifier. Next, we will show some techniques that can be applied to extract
features from text data.
1.6 Evaluation
The experimental evaluation of text classifiers measures efficiency (i.e.,
capacity to execute the correct classification). Accuracy, recall, and preci-
sion are generally employed to measure the effectiveness of a text classifier.
Accuracy (FP+FN/TP+TN+FP+FN = 1-accuracy), on the other hand, is
not w idely used for text classification applications because it is insensitive
to variations in the number of correct decisions due to the large value of
the denominator (TP + TN), as discussed below in Table 1.4.
dataset, the third phase uses TF-IDF feature extraction techniques, the
fourth phase is to train and test our model (after dividing the dataset into
two parts: 70% training set and 30% testing set to fit the model), and the
last phase is to predict and evaluate the model by using amusement preci-
sion, recall, accuracy, and f score confusion matrix, as shown in Table 1.5
and Figure 1.5 below.
Table 1.5 shows the outcome of three classifier models. Through accu-
racy, recall, precision, and F1-Measure, Naïve Bayes achieved the higher
accuracy at around 97.06% among classifier models for detection of spam
files from a data set.
As shown in the Figure 1.5, Naïve Bayes achieved higher results and
higher improvement accuracy with a data trained spam model using
K-Nearest Neighbor and Logistic Regression algorithms.
98
97
96 Naïve Bayes
95 Knearest Neighbor
94
Logistic Regression
93
92
91
90
89
88
Accuracy (%) Recall (%) Precision (%) F1‐Measure
Risk
Management
Fraud Customer
Detection Applications Care Services
of Text
Mining
1.9.3 Law
Government institutions have generated large volumes of legal record text
documents. Analysis and retrieving that data manually is so difficult. Here,
a system is required to deal with information and process it automatically
in order to help lawyers and their clients [22]. Organization of those record
text documents is the foremost difficulty to the law community. Building a
system to classify documents is helpful for the law community.
1.10 Conclusions
In this review paper, we present a review of text data mining techniques with
relevant descriptions. Text data mining techniques are applied to extract
readable knowledge from raw text data sets regularly into the unstructured
data. So, in this paper we presented the gist of various previous research
works presented on the predominant text mining techniques, namely Data
Mining, Text Classification, Information Extraction, Question Answering,
Topic Tracking, Natural Language Processing, Information Retrieval,
Text Summarization, Text Categorization, Clustering, and Information
Visualization. The paper has presented the important mechanism of
preprocessing text data mining with four steps: Tokenization, Filtering,
Lemmatization, and Stemming. We have discussed the process of text data
mining by presenting the fields that use text mining applications. Lastly,
we compared the most popular text classification algorithms. Finally, we
used a spam dataset for classification detection data and a three classifier
algorithm with TF-IDF feature extraction, achieving higher accuracy with
Naïve Bayes around 97.06%. Illustrations of the usage of text classification
as support for applications in medicine, law, education, etc. are related in
a separate part.
Review on Text Classification and TM Techniques 15
References
1. S. S. Kermode and V. B. Bhagat, “A Review: Detection and Blocking Social
Media Malicious Posts,” Int. J. Mod. Trends Eng. Res., vol. 3, no. 11, pp. 130–
136, 2016, doi: 10.21884/IJMTER.2016.3133.Q4M8O.
2. M. Allahyari et al., “A Brief Survey of Text Mining: Classification, Clustering
and Extraction Techniques,” 2017, [Online]. Available: http://arxiv.org/
abs/1707.02919.
3. C. C. Aggarwal and C. X. Zhai, “Mining text data,” Min. Text Data, vol.
9781461432, pp. 1–522, 2013, doi: 10.1007/978-1-4614-3223-4.
4. M. A. Zende, M. B. Tuplondhe, S. B. Walunj, and S. V. Parulekar, “Text min-
ing using python,” no. 3, pp. 2393–2395, 2016.
5. N. F. F. Da Silva, E. R. Hruschka, and E. R. Hruschka, “Tweet sentiment
analysis with classifier ensembles,” Decis. Support Syst., vol. 66, pp. 170–179,
2014, doi: 10.1016/j.dss.2014.07.003.
6. A. Nielsen, “Python programming — text and web mining,” Aust. J. Plant
Physiol., vol. 21, pp. 507–516, 2011, [Online]. Available: http://www2.imm.
dtu.dk/pubdb/views/edoc_download.php/5781/pdf/imm5781.pdf.
7. C. W. Lee, S. A. Licorish, B. T. R. Savarimuthu, and S. G. Macdonell,
“Augmenting text mining approaches with social network analysis to under-
stand the complex relationships among users’ requests: A case study of the
android operating system,” Proc. Annu. Hawaii Int. Conf. Syst. Sci., vol.
2016-March, pp. 1144–1153, 2016, doi: 10.1109/HICSS.2016.145.
8. N. U. Pannala, C. P. Nawarathna, J. T. K. Jayakody, L. Rupasinghe, and K.
Krishnadeva, “Supervised Learning-Based Approach to Aspect Based
Sentiment Analysis,” 2016 IEEE Int. Conf. Comput. Inf. Technol., pp. 662–
666, 2016, doi: 10.1109/CIT.2016.107.
9. A. Swapna, K. G. Guptha, and K. Geetha, “Efficient Approach for Web Search
Personalization in User Behavior Supported Web Server Log Files Using Web
Usage Mining,” vol. 2, no. 4, pp. 189–196,
10. G. Barbier, “Social Network Data Analytics,” 2011, doi: 10.1007/978-1-
4419-8462-3.
11. T. Singh, M. Kumari, T. L. Pal, and A. Chauhan, “Current Trends in Text
Mining for Social Media,” Int. J. Grid Distrib. Comput., vol. 10, no. 6,
pp. 11–28, 2017, doi: 10.14257/ijgdc.2017.10.6.02.
12. A. Shahrestani, M. Feily, R. Ahmad, and S. Ramadass, “Architecture for
applying data mining and visualization on network flow for botnet traffic
detection,” ICCTD 2009 - 2009 Int. Conf. Comput. Technol. Dev., vol. 1,
pp. 33–37, 2009, doi: 10.1109/ICCTD.2009.82.
13. L. Kumar and P. K. Bhatia, “Available Online at www.jgrcs.info TEXT
MINING: CONCEPTS, PROCESS, AND APPLICATIONS,” vol. 4, no. 3,
pp. 36–39, 2013.
16 Mathematics and Computer Science Volume 2
Abstract
The rotational turbulence caused by mixing the layers of air, wind shear compo-
nents, mountain waves, aerosol particles, and other pollutants affects the lowest
and densest layer of the earth’s surface troposphere. Due to the turbulence, the
height of the convective boundary layer (CBLH) changes over the day dramat-
ically. We observe the variation in peak positions of lidar backscatter signals by
performing a statistical technique for analyzing the behavior of the convective
boundary layer (CBL). After that, to examine the behavior of the whole boundary
layer, a distribution method and histogram plots will be used. We provide the sta-
tistical method for getting the best fit distribution to show how the result leads to
the physical observation of data.
2.1 Introduction
Surface forcing of the planetary boundary layer influences the closest and
deepest part of the earth’s atmosphere, i.e., the troposphere. The turbu-
lence in the troposphere causes the high variability of the temperature,
moisture, wind shear, and pollutant particles. Due to surface friction,
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (19–32) © 2023 Scrivener Publishing LLC
19
20 Mathematics and Computer Science Volume 2
winds in ABL are weaker than above and tend to blow towards the area
where pressure is low. It is noticeable over the Indian subcontinent that the
height of the convective boundary layer (CBLH) is low during the winter
and monsoon and high in the sunny days as the warmer lower-layer air
mixes with the cool air and convection arises in the summertime only.
This behavior indicates that solar heating affects the CBLH, and hence,
CBL is also named the daytime planetary boundary layer. The convec-
tion occurs by mixing of airs forming bubbles of warmer air or eddies.
The small eddies generated by the local wind shear cause turbulence in
the surface layer and have a completely random behavior, however, the
turbulence in the mixed layer is caused by large bubbles of warmer air,
and hence, it is not completely random. Since the turbulent processes are
found to be nondeterministic, it is beneficial to use a robust technique
like statistical methods to find out the variability in the boundary layer
and its key parameter as CBLH and has been studied by Pal, S., Behrendt,
A., and Wulfmeyer V. in Elastic-backscatter-lidar-based characterization
of the convective boundary layer and investigation of related statistics [3].
The signal which will be used in the study of CBLH is the total backscat-
ter from the atmosphere that represents combined aerosol and molecular
backscatter as mentioned earlier by Dang R, Yang Y, Hu X-M, Wang Z,
and Zhang S [2]. For using the statistical method, we need the informa-
tion in terms of data and we choose an instrument with high resolution
power named Lidar (Light Detection and Ranging), which is a remote
sensing method used in measuring the ranges to the earth and the concept
of using Lidar to detect ABL height relies on the assumption that there is
a strong gradient in the concentration of aerosols in the CBL versus the
free atmosphere. The importance of defining a new technique is that this
method works for single scan data and uses spatial average, however, the
variance method needs multiple numbers of scans to identify the variance
and use time average.
After finding the CBLH through statistical technique, we will try to
examine the behavior of the whole boundary layer rather than just the CBL
part. The histogram plots and distribution method can serve this purpose
very well, meaning a list of best fit distribution along with each signal can
be used to examine the behavior of the boundary layer, used before by the
authors in ‘The Determination of Aerosol Distribution by a No-Blind-Zone
Scanning Lidar” [6]. After getting the list of all possible distributions, it will
be checked whether there is any range of the signals which follow a speci-
fied distribution or if the signals are following the distributions randomly.
Now, to find the distribution of the backscatter signals manually for any
structure of the boundary layer, a new technique needs to be introduced
Study of Lidar Signals of the ABL Using Statistical Technique 21
which holds in practical life. In this new technique, we will try to use the
statistical approach. Lastly, we check the consistency for applying both the
statistical techniques by using two different datasets and relate the math-
ematical result with the physical observations to examine the behavior of
ABL.
2.2 Methodology
2.2.1 A Statistical Approach to Determine the CBLH
Due to the high-resolution power of Lidar, signals of Lidar can be used in
finding the CBLH as Georgoulias, A.K., Papanastasiou, D.K., and Melas, D.
et al. used in “Statistical analysis of boundary layer heights in a suburban
environment” [4]. There are different types of Lidar techniques that serve a
specific purpose, such as temperature profile using vibrational Raman (VR)
and rotational Raman (RR). Backscattering Lidar and Elastic-backscatter
Lidar (EBL) were used in ‘New Technique to Retrieve Tropospheric
Temperature Using Vibrational and Rotational Raman Backscattering’ [5].
In our statistical technique, we do simple differentiation of the signal con-
cerning the height, i.e., tracking the local maxima points, then we do some
restrictions on the points of local maxima to indicate the real position of
peaks of backscatter signal. The peak position changes with change in val-
ues of the restriction set. The peak position changes with change in values
of the restriction set given below in Table 2.1.
Table 2.1 Restriction set used in algorithm for detecting peak position of lidar
signals.
Elements of set of
restriction Meaning Default value
mph Minimum peak height None
mpd Minimum distance between 1 (positive integer)
two peaks
threshold Noise level 0
kpsh Keep peaks with the same False
height even if they are closer
than `mpd`
valley Local minima False
22 Mathematics and Computer Science Volume 2
EN = n(i)/N (2.1)
where n(i) is the number of points that are less than Xi.
K-S Test:
(Null hypothesis) H0: data follows given distribution
(Alternative hypothesis) H1: data does not follow the given dis-
tribution function
24 Mathematics and Computer Science Volume 2
Test Statistic:
This test is defined as:
i −1 i
=D max F ( Xi ) − , − F ( Xi ) (2.2)
1≤i ≤N N N
×106
3 9
8
2.5 7
Range (km)
6
2
5
1.5 4
3
1 2
1
0.5
0
09:24 09:25 09:26 09:27 09:28 09:29 09:30 09:31 09:32 09:33
Time (IST)
the same figures for the larger time period from 13:00 to 19:00 UTC and
from 17:00 to 21:00 UTC [1]. The reason for taking this time interval is
that the turbulence in the CBL increases with the rise in sunrays, i.e., in
the morning and due to the temperature variation boundary layer thermal
increases. However, if we take the time-interval in the night, then turbu-
lence ceases with the decrease in the amount of solar heating.
After using the statistical technique for all the backscatter signals, it is
observed that after a certain height of signal, data behaves similarly for all
the backscatter signals, which means the point at which there is a steep
gradient approximately has the same location for each signal. This means
that after a certain height of the layer, less presence of aerosols and other
pollutant particle ceases the turbulence. Hence, a part of the backscatter
signal in which the turbulence is very high is used rather than using the
whole range. After getting the peak points for each scan, we make a graph
of peak points along with scan number for a time-interval.
Now, if we compare these two figures, then it can be seen that the graph
of peak points detected using the statistical technique matches the figure
containing the rise of boundary layer thermals. When one observes graphs
of the peak points for each signal in the included excel sheet, it can be eas-
ily seen that in starting there is a peak point near the first peak and after
some signals, there is no point of the peak near the first peak point. Again,
there is a peak point found near the first one which shows the dissimilarity
in the location of peaks of the backscatter signals. From Figure 2.2, one can
notice that peak points are randomly distributed over the interval. Also,
the changes in peak height are not following any particular pattern. This
result shows that the changes in the part of the surface layer of CBL do
not follow any identified pattern and hence cannot be organized into any
predefined structure.
26 Mathematics and Computer Science Volume 2
2000
Altitude (m)
1500
0924 0925 0926 0927 0928 0929 0930 0931 0932 0933
Local Time (LT)
However, after excluding a certain range, peak points attain the approx-
imately same position, which physically shows that the turbulence above
this shallow layer in the mixed layer is not completely random and can be
organized into some predefined structure like plums and thermals. This
method seems to be consistent and to confirm this we will take another
dataset with the time interval 09:34-09:44, having the same interval length.
The same process will be used here and after getting the peak points for
each scan, we will make a graph of peak points for dataset-2 along with
scan number. This method seems to be consistent and to confirm this we
will take another dataset with the time interval 09:34-09:44, having the
same interval length as shown in Figure 2.3. The same process will be used
here and after getting the peak points for each scan, we will make a graph
of peak points for dataset-2 along with the scan number as shown in Figure
2.4.
×106
3 9
8
2.5 7
6
Range (km)
2
5
1.5 4
3
1 2
1
0.5
0
09:34 09:35 09:36 09:37 09:38 09:39 09:40 09:41 09:42 09:43 09:44
Time (IST)
2000
Altitude (m)
1500
0934 0935 0936 0937 0938 0939 0940 0941 0942 0943
LOCAL TIME (LT)
In this, if we compare both the figures then the peak points clearly show
the temporal variation of Lidar signals. Hence, the result shows the consis-
tency of the statistical method used here and this method can be utilized in
detecting peak positions of backscatter signals.
After using the statistical technique (k-s test statistic), the distributions
which fit the dataset of the backscatter signals have been found and a list of
them along with the histogram plot, as well as the graph of best-fit distribu-
tion in an excel file are included in the link. Now, if we analyze the graphs
up to the two-thirds part of the dataset, gamma or beta distribution fits in
its maximum part, which has an almost same graph, and in the remaining
one-third of the dataset, t-distribution fits perfectly. However, a few data-
sets perform the log norm, chi-square, inverse Gaussian, and Pearson 3
distributions. If we use the same technique for dataset-2 and analyze the
graphs, the whole dataset follows t-distribution. This is probably happen-
ing because we have taken this dataset for the time-interval which is after
the time for the first dataset.
that the changes in the distribution with some particular ranges of signals
are happening due to the physical process involved in that. Sometimes we
get a broader spectrum and sometimes a narrower spectrum. The reason
for this change is that the broader spectrum indicates turbulence process
and wind variability, whereas the narrower spectrum indicates tempera-
ture inversion and is associated with stable layer phenomenon. The fitting
of distribution is affected by thermal heating, water vapor condensation,
and other physical processes involved during the day. So, we have exam-
ined the natural environment and now this study leads us to examine the
changes in the manmade environment. The consistency of this method is
paving the way to use it in different fields such as hydrodynamic modeling,
storm surge modeling, etc.
Acknowledgement
I would like to express my gratitude to the team of Lidar project of NARL,
Tirupati, India for providing the data for the research work.
References
1. de Arruda Moreira, G., da Silva Lopes, F. J., Guerrero-Rascado, J. L., da
Silva, J. J., Arleques Gomes, A., Landulfo, E., and Alados-Arboledas, L.,
07 Aug 2019. Analyzing the atmospheric boundary layer using high-or-
der moments obtained from multi- wavelength lidar data: impact of wave-
length choice, Atmos. Meas. Tech., 12, 4261– 4276, https://doi.org/10.5194/
amt-12-4261-2019.
2. Dang R, Yang Y, Hu X-M, Wang Z, Zhang S, 2019. A Review of Techniques
for Diagnosing the Atmospheric Boundary Layer Height (ABLH) Using
Aerosol Lidar Data. Remote Sensing. 11(13):1590. https://doi.org/10.3390/
rs11131590.
3. Pal, S., Behrendt, A., and Wulfmeyer, V., 2010. Elastic-backscatter-lidar-
based characterization of the convective boundary layer and investigation
of related statistics, Ann. Geophys., 28, 825–847, https://doi.org/10.5194/
angeo-28-825-2010.
4. Georgoulias, A.K., Papanastasiou, D.K., Melas, D. et al., 2009. Statistical anal-
ysis of boundary layer heights in a suburban environment. Meteorol Atmos
Phys 104, 103– 111. https://doi.org/10.1007/s00703-009-0021-z.
5. Su, J., McCormick, M. P., & Lei, L. (2020). New Technique to Retrieve Tropo
spheric Temperature Using Vibrational and Rotational Raman Backscattering.
Earth and Space Science, 7. https://doi. org/10.1029/2019EA000817.
Study of Lidar Signals of the ABL Using Statistical Technique 29
6. Wang, J.; Liu, W.; Liu, C.; Zhang, T.; Liu, J.; Chen, Z.; Xiang, Y.; Meng, X, 2020.
The Determination of Aerosol Distribution by a No-Blind-Zone Scanning
Lidar. Remote Sens. 12, 626. https://doi.org/10.3390/rs12040626
7. QingWeiPing, 2011. Analysis of ocean clutter for wide-band radar based on
real data https://doi.org/10.1145/2071639.2071669
8. Yuan, J.; Lv, X.; Li, R, 2018. A Speckle Filtering Method Based on Hypothesis
Testing for Time-Series SAR Images. Remote Sensing, 10, 1383. https://doi.
org/10.3390/rs10091383.
9. Hunter MO, Keller M, Morton D, Cook B, Lefsky M, Ducey M, et al. (2015)
Structural Dynamics of Tropical Moist Forest Gaps. PLoS ONE 10(7):
e0132144. https://doi.org/10.1371/journal.pone.0132144
10. J. Luo, S. Wang, E. Zhang and J. Luo, 2015. “Non-cooperative signal detection
in alpha stable noise via Kolmogorov-Smirnov test” 2015 8th International
Congress on Image and Signal Processing (CISP), 2015, pp. 1464-1468, doi:
10.1109/CISP.2015.7408114.
30 Mathematics and Computer Science Volume 2
Annexure
Flowchart and Algorithm for peak detection
Flowchart:
START
No
for j=1,
j≤=n,
Yes
No
for i=1,
j<size,
No
Yes
No Track if value
maxim >mph
Yes
ZK = value
Yes
ZK+1 - ZK
≥mpd
No
Yes
Algorithm:
Step-1. Start
Step-2. Input the minimum peak height (mph) and minimum distance
between two peaks (mpd).
Step-3. Start a for loop to read columns of multidimensional data.
Step-4. Read the data of the column if ‘for’ loop is true else, go to step 11.
Step-5. Introduce a variable x to store the data in an array form. Then, use
the size function to find the size of x and after that, use the differentiation
function to differentiate the variable x.
Step-6. Again, start a ‘for’ loop for data stored in x.
Step-7. Track local maxima of data if for loop is true else, go to step 4 and
restart the process for the next column.
Step-8. If it is a point of maxima then, check whether its value is greater
than mph or not. If yes, go to step otherwise go to step 7.
Step-9. Introduce a new variable Zk and store the value in it.
Step-10. If difference between two local maxima i.e. Zk+1 - Zk is greater than
mpd then, assign this value as peak value else go to step 7.
Step-11. Exit.
Abstract
In this paper, a new stochastic framework to determine optimal combination thera-
pies in colon cancer-induced immune response is presented. The dynamics of colon
cancer are described through an It¨o stochastic process, whose probability density
function evolution is governed by the Fokker-Planck equation. An open-loop control
optimization problem is pro- posed to determine the optimal combination thera-
pies. Numerical results with combination therapies comprising of the chemotherapy
drug Doxorubicin and immunotherapy drug IL-2 validate the proposed framework.
3.1 Introduction
Colon cancer is a leading cause of global cancer related deaths [11]. The lack
of early symptoms forces the detection of colon cancer to take place at the
metastatic phase of the cancer [29]. Thus, it becomes important to devise fast
and accurate treatment strategies for a cure. In this context, combination ther-
apies have been clinically shown to be an effective strategy for combating can-
cer in comparison to monotherapy (see [18]). Conventional monotherapeutic
techniques are indiscriminate in choosing actively growing cells that lead to
the death of not only cancerous cells but also healthy cells. E.g., chemotherapy
drugs can be toxic and lead to multiple side effects and risks by weakening the
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (33–48) © 2023 Scrivener Publishing LLC
33
34 Mathematics and Computer Science Volume 2
patient’s immune system, targeting the bone-marrow cells [22]. This leads to
increased susceptibility to secondary infections. However, with combination
therapies, even though it might still be toxic if there is a presence of a chemo-
therapy drug, the toxicity effect is significantly diminished due to different tar-
gets being affected. Moreover, since the effect of combination therapies work
in synergy, lower dosages of the individual drugs in the combination therapy
are required for treatment, which further reduces the toxic effects [19].
To develop optimal combination therapies in colon cancer, it is important
to understand relevant biomarkers that govern the progression of the cancer
[8]. One such biomarker is the immune response to the cancerous cells. It
has been observed that onsite immune reactions at the tumor locations and
prognosis are highly correlated and independent of the size of the tumor
[12]. Furthermore, a high expression of various immune pathways like Th1
and Th17 are associated with a poor prognosis or prolonged disease-free
survival for patients with colon cancer [32]. Thus, it becomes important to
develop optimal personalized treatments for colon cancer patients, taking
into account the various types of immune cells, their numbers, and interac-
tion with the cancer cells and each other. Since drugs for treatment of colon
cancer like chemotherapeutic drugs and immunotherapies are quite costly,
it is expensive to perform in-vitro and vivo experimental studies for testing
effects of such drugs. An alternative cost effective option is to develop com-
putational frameworks for testing optimal combination drug dosages [7].
This can be done using pharmacokinetic models that are given by a set of
ordinary or partial differential equations (ODE or PDE).
There are several dynamic models that have been developed to describe the
immune response and interactions in colon cancer and associated therapies.
In [4], a combined compart- mental model with that of irinotecan is used to
describe the physiology of colon cancer. A set of ODEs were used to represent a
3D structure of colon cancer in [14]. In [15], the authors use describe the phe-
nomenon of tumorigenesis in colon cancer using mathematical modeling. The
authors in [17], describe the initiation of colon cancer and its link with colitis
using a mathematical framework. In [23], the authors use a pharmacokinetic
cellular automata model to incorporate the cytotoxic effects of chemotherapy
drugs. In [13], dynamical systems are used to describe multiple pathways in
colon cancer. For a detailed review of various mathematical models for colon
cancer-induced immune response, we refer the reader to [3].
We contribute to the field of pharmacokinetic cancer research by present-
ing an effective approach to develop personalized therapies for colon cancer-
induced immune response. The starting point of this estimation process is the
dynamic model for colon cancer induced immune response, given in [10].
The model describes the evolution of four variables: the tumor cell count,
concentration of the Natural Killer cells, concentration of the CD8+ cells, and
Therapies in Colon Cancer using FP Framework 35
dT
= aT − abT 2 − DT − cNT − α1u1 (t )T , T (0) = T0 ,
dτ
dN
= eC − fN − pNT − α 2u1(t )N + β2u2 (t )N , N (0) = N0 ,
dτ
dL T
= mL + j L − qLT + (r1N + r2C )T − α 3u1(t )L
dτ k +T
+ β3u2 (t )L, L(0) = L0 ,
dC
= α − β C − α 4u1(t )C, C(0) =C0 ,
dτ
(3.1)
1
where D = d ,T0 , N 0 , C0 represent the initial conditions for T, N,
s(T /L )l + 1
L, and C, respectively, and u1, u2 represent dosages of Doxorubicin and
IL-2, respectively.
The parameters of the system (3.1) are defined in [10]. To stabilize the
solutions of the numerical methods, we non-dimensionalize the ODE
system (3.1) using the following non-dimensionalized state variables and
parameters:
=T k=
1T , N k2=
N , L k=
3 L, C k4C , t = k5τ ,
a b c d ek f 1010 p
=a= ,b = ,c = ,d , e = 2= ,f = ,p ,
k5 k1 k2k5 k5 k4 k5 k5 k1k5
m j 108 q r1k3 r2k3
=m = ,j = , k k= 1k , q = , r1 = = , r2 ,
k5 k5 k5k1 k1k2k5 k1k4 k5
α k4 β αi βi
= =
s 250 s, α =,β = l l, α= i = ,. . ., 4, βi
, i 1= i = 1, 2.
k5 k5 k5 k5
(3.2)
Therapies in Colon Cancer using FP Framework 37
dT
= aT − abT 2 − DT − cNT − α1u1(t )T , T (0) = T0
dt
dN
= eC − fN − 10−10 pNT − α 2u1(t )N + β2 u2 (t )N , = N (0) = N0 ,
dt
dL T
= mL + j L − 10−8 qLT + ( r1 N + r2 C)T − α 3 u1(t) L
dt k +T
+ β 3 u 2(t) L , L (0) =L0 ,
dC
= α − β C − α 4 u1 (t )C , C (0) =
C0 , (3.3)
dt
(L / T )ll
where D = d .
4 s ⋅10 ⋅ (k1 / k3 ) + (L / T )
The compact form of the aforementioned system of ODEs, given in
(3.3), is as follows:
dX
= F ( X , U ),
dt
X (0) = X 0 , (3.4)
T
where X (t ) = (T (t ), N (t ), L (t ), C(t )) .
We extend the ODE system (3.3) to include stochasticity present in
the dynamics. For this purpose, we consider the Itô stochastic differential
equation corresponding to (3.3).
where
1
∂ t f (x , t ) + ∇ ⋅ (F (x , U ) f (x , t ))= ∇ ⋅ (σ 2 (x )∇f (x , t )),
2
f (x , 0) = f 0 (x ), (3.7)
σ 2j
H j (x , t ; f=
) ∂ x j f − Fj (x , U) f , j =
1, 2, 3, 4.
2 (3.9)
H ⋅n =0 on ∂Ω × (0, T f ),
^
(3.10)
α ν T ν T
=U * arg
= minU∈Uad J ( f , U) :
2Q∫( f (x , t ) − f * (x , t ))2 dx + 1 ∫ u1 (t ) 2 dt + 2 ∫ u2 (t ) 2 dt ,
20 2 0
(3.11)
subject to the FP system (3.7),(3.10), where the desired PDF is f *(x, t).
∂
|| f (t )||2 2 = − ||σ∇f (t )||2 2 + 2Ω (F (t )) ⋅ σ −1σ∇f (t )dx. (3.13)
∂t L ( Ω) L (Ω)
The last term in (3.13) can be estimated using the Young’s inequality,
2bd ≤ kb2 + d2/k, with k = || σ−1 ||2, which is the L2 matrix norm of σ−1. We
then obtain the following:
∂
|| f (t )||22 ≤ ||σ −1 ||2 N 2 || f (t )||22
∂t L (Ω) 2 L (Ω)
Theorem 3.1. Let f0 ∈ H1(Ω) and let J be given as in (3.11). Then, there exists
(f ∗, U ∗) ∈ C([0, Tf ]; H1(Ω)) × Uad with f ∗ being a solution to E(f0, U ∗) = 0
and U ∗ minimizing J in Uad.
Proof. Since J is bounded, a minimizing sequence (Um) exists in Uad.
Moreover, J is being coercive and sequentially weakens lower semi-continu-
ous in Uad, implying the boundedness of this sequence. Due to the fact that Uad
is a closed and convex subset of a Hilbert space, the sequence (Um) contains a
convergent subsequence (U ml) in Uad, such that U ml → U ∗. Correspondingly,
the sequences (f ml) and (∂t f ml), where fml = Λ(θml) are bounded in L2([0, Tf ];
H1(Ω)), L2([0, Tf ]; H−1(Ω)), respectively. This implies the weak convergence
of the sequences to f ∗ and ∂t f ∗, respectively.
Therapies in Colon Cancer using FP Framework 41
1
∂ t f ( x , t ) + ∇ ⋅ (F ( x , U ) f (x , t ))= ∇ ⋅ (σ 2∇f (x , t )), in Ω × (0, Tf ),
2
f (x , 0) = f 0 ( x ), in Ω, (FOR )
H ⋅n =
^
0, on ∂Ω × (0, Tf ).
1
−∂t p(x , t ) − f (x , t )(F (x ,θ ) ⋅∇p(x , t )) − ∇ ⋅ (σ 2∇p(x , t ))
2
*
= −α ( f (x , t ) − fi (x , t )), in Ω × (0,T f ),
p(x ,T f ) = 0, in Ω,
∂p
= 0, on ∂Ω × (0,T f ). (ADJ)
∂n
∫
βU − ∇U F ⋅∇p,ψ − U
Ω
L2 ([0,T ])
≥ 0, ∀ψ ∈U ad . (OPT)
Ωh = {(x1, x2, x3, x4) ∈ R4 : (x1i, x2j, x3k, x4l) = (x10 + ih, x20 + jh, x30 + kh,
x40 + lh)},
Qh,δt = {(x1i, x2j, x3k, x4l, tm) : (x1i, x2j, x3k, x4l) ∈ Ωh, tm = mδt, 0 ≤ m ≤ Nt}.
3.5 Results
This section describes the numerical results for obtaining optimal treat-
ment strategies with the aforementioned FP framework. We choose our
domain Ω = (0, 6)4 and discretize it using N xi = 51 points for i = 1, 2, 3, 4.
The final time Tf is chosen to be 10 and the maximum number of time
steps Nt is chosen to be 200. The values of the constants used in converting
the ODE system (3.1) to its non-dimensional form given in (3.3) are given
as k1 = 10−10, k2 = 10−5, k3 = 10−7, k4 = 10−8, and k5 = 1. To obtain the target
PDF f ∗(x, t), we simulate the ODE system (3.3) with the u1 , u2 set to 0 and
with the value of the non-dimensional parameters (d , l , s ) = (2.1,1.1,1.25)
that represents a patient with strong immune response [10]. The values
of the other parameters are taken from [10]. After we obtain the trajec-
tories of T¯, N¯, L¯, and C, we then choose 20 time points ti, and at each
point (T¯(ti), N¯ (ti), L¯(ti), C¯(ti)) assign a Gaussian PDF given by a normal
distribution with variance 0.05. We finally perform a 5D interpolation to
obtain the desired PDF f ∗(x, t). Based on a statistical analysis of the dataset
given in [10], we choose
×1010 Plot of T (without treatment) ×109 Plot of T (with treatment) Plot of u1 ×105 Plot of u2
4 2 4 2
3.5 1.8 3.5 1.8
1.6 1.6
3 3
1.4 1.4
2.5
u2 (IUIL-2/l)
2.5
u1 (mg)
1.2 1.2
T(cells)
T(cells)
2 1 2
1
1.5 0.8 1.5
0.8
0.6
1 1 0.6
0.4
0.5 0.2 0.5 0.4
0 0 0 0.2
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
t (days) t (days) t (days) t (days)
(a) T (without treatment) (b) T (with treatment) (c) Doxorubicin profile (d) IL-2 profile
Figure 3.1 Test case 1: plots of mean tumor profiles and drug dosages over 10 day period.
×1010 Plot of T (without treatment) ×109 Plot of T (with treatment) Plot of u1 ×104 Plot of u2
3.5 2 3 10
1.8 9
3 2.5
1.6 8
2.5 1.4 7
2
u2 (IUIL-2/l)
1.2 6
u1 (mg)
2
T(cells)
T(cells)
1 1.5 5
1.5 0.8 4
1 3
1 0.6
0.4 0.5 2
0.5 0.2 1
0 0 0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
t (days) t (days) t (days) t (days)
(a) T (without treatment) (b) T (with treatment) (c) Doxorubicin (d) IL-2
Figure 3.2 Test case 2: plots of mean tumor profiles and drug dosages over 10 day period.
is far less than 70 mg. Moreover, the observed daily dosage of IL-2 from
Figure 3.1 is less than the standard dosage of IL-2 over a day, which is 2.1 ×
106 IU IL-2/l. This suggests that optimal combination therapies adminis-
tered daily leads to lower total dosages and, thus, lower toxicity effects.
In Test Case 2, we choose the same initial values of T , N , L , and C, along
with the values of the non-dimensional parameters (d , l , s ) = (1.6 , 1.4 , 2)
that represent patients with moderately strong immune systems.
Figure 3.2 shows the plots of the tumor profiles without and with treat-
ment and the optimal daily dosages of Doxorubicin and IL-2. We observe
a similar behavior as Test Case 1. In addition, we note that the dosages
are smaller compared to the patient in Test Case 1 since the immune sys-
tem is moderately strong. This shows the effectiveness of our framework in
obtaining optimal dosages in colon cancer.
3.6 Conclusion
In this paper, we presented a new stochastic frameworks to determine opti-
mal combination therapies in colon cancer-induced immune response. We
Therapies in Colon Cancer using FP Framework 45
Acknowledgments
The authors were supported by the National Institutes of Health (Grant
Number: R21CA242933).
S. Roy was also partially supported by the Interdisciplinary Research
Program, University of Texas at Arlington, Grant number: 2021-772. The
content is solely the responsibility of the authors and does not necessarily
represent the official views of the National Institutes of Health.
References
1. M. Annunziato and A. Borzì. A fokker–planck control framework for mul-
tidimensional stochastic processes. Journal of Computational and Applied
Mathematics, 237(1):487–507, 2013.
2. M. Annunziato and A. Borzì. A fokker–planck approach to the reconstruc-
tion of a cell membrane potential. SIAM Journal on Scientific Computing,
43(3):B623–B649, 2021.
3. A. Ballesta and J. Clairambault. Physiologically based mathematical mod-
els to optimize therapies against metastatic colorectal cancer: a mini-review.
Current pharmaceutical design, 20(1):37–48, 2014.
4. A. Ballesta, J. Clairambault, S. Dulong, and F. Levi. A systems biomedicine
approach for chronotherapeutics optimization: focus on the anticancer drug
irinotecan. In New Challenges for Cancer Systems Biomedicine, pages 301–
327. Springer, 2012.
5. A. Borzì and L. Grüne. Towards a solution of mean-field control problems
using model predictive control. IFAC-PapersOnLine, 53(2):4973–4978, 2020.
6. J. Chang and G. Cooper. A practical difference scheme for fokker-planck
equations. Journal of Computational Physics, 6(1):1–16, 1970.
7. Y. Chang, M. Funk, S. Roy, E. Stephenson, S. Choi, H. V. Kojouharov, B.
Chen, and Z. Pan. Developing a mathematical model of intracellular calcium
46 Mathematics and Computer Science Volume 2
20. S. Pal and S. Roy. A new non-linear conjugate gradient algorithm for
destructive cure rate model and a simulation study: illustration with nega-
tive binomial competing risks. Communications in Statistics-Simulation and
Computation, pages 1–15, 2020.
21. S. Pal and S. Roy. On the estimation of destructive cure rate model: A
new study with exponentially weighted poisson competing risks. Statistica
Neerlandica, 75(3):324–342, 2021.
22. A. H. Partridge, H. J. Burstein, and E. P. Winer. Side effects of chemotherapy
and combined chemohormonal therapy in women with early-stage breast
cancer. JNCI Monographs, 2001(30):135–142, 2001.
23. G. G. Powathil, D. J. Adamson, and M. A. Chaplain. Towards predicting the
response of a solid tumour to chemotherapy and radiotherapy treatments:
clinical insights from a computational model. PLoS computational biology,
9(7):e1003120, 2013.
24. S. Roy, M. Annunziato, and A. Borzì. A fokker–planck feedback control-
constrained approach for modelling crowd motion. Journal of Computational
and Theoretical Transport, 45(6):442–458, 2016.
25. S. Roy, M. Annunziato, A. Borzì, and C. Klingenberg. A fokker–planck
approach to control collective motion. Computational Optimization and
Applications, 69(2):423–459, 2018.
26. S. Roy, A. Borzì, and A. Habbal. Pedestrian motion constrained by fp-
constrained nash games. Royal Society Open Science, 4:170648, 2017.
27. S. Roy, S. Pal, A. Manoj, S. Kakarla, J. Villegas, and M. Alajmi. A fokker-planck
framework for parameter estimation and sensitivity analysis in colon cancer.
AIP Conference Proceedings (accepted, in press), 2522, 070005, 2022.
28. S. Roy, Z. Pan, and S. Pal. A fokker-planck feedback control framework
for optimal personalized therapies in colon cancer-induced angiogenesis.
Journal of Mathematical Biology, 84:23, 2022.
29. H. Schmoll, E. Van Cutsem, A. Stein, V. Valentini, B. Glimelius, K.
Haustermans, B. Nordlinger, C. Van de Velde, J. Balmana, J. Regula, et al.
Esmo consensus guidelines for management of patients with colon and rectal
cancer. a personalized approach to clinical decision making. Annals of oncol-
ogy, 23(10):2479–2516, 2012.
30. T. Tao. Nonlinear dispersive equations: local and global analysis. Number 106.
American Mathematical Soc., 2006.
31. V. Thalhofer, M. Annunziato, and A. Borzì. Stochastic modelling and control
of antibiotic subtilin production. Journal of mathematical biology, 73(3):727–
749, 2016.
32. M. Tosolini, A. Kirilovsky, B. Mlecnik, T. Fredriksen, S. Mauger, G. Bindea,
A. Berger, P. Bruneval, W.-H. Fridman, F. Pagès, et al. Clinical impact of dif-
ferent classes of infiltrating t cytotoxic and helper cells (th1, th2, treg, th17)
in patients with colorectal cancer. Cancer research, 71(4):1263–1271, 2011.
4
Detection and Classification of
Leaf Blast Disease using Decision
Tree Algorithm in Rice Crop
Sarvesh Vishwakarma1* and Bhavna Chilwal2
2
Department of CSE, DIT University Dehradun, Uttarakhand, India
Abstract
The agricultural field is the most important field for any nation but some issues pre-
vail and affect agricultural products every year. Agricultural diseases are the main
concern for yield loss. This chapter uses the Decision Tree technique to form a tree
structure for leaf blast disease level detection in rice crops. A Decision Tree is used
as a classification technique and here disease levels are classified based on symp-
toms that occur during infection. The Iterative Dichotomiser 3 (ID3) algorithm
is one of the important methods to form a Decision Tree based on entropy and
information gain. The nodes in the tree are the symptoms that have different labels
for disease occurrence. This decision tree will help detect the occurrence of disease
as per the symptoms and help farmers get information about the severity level of a
disease so that they can take required measures on time to save the crop from loss.
Keywords: Decision tree, ID3 algorithm, entropy, information gain, leaf blast
disease
4.1 Introduction
There are so many fungal diseases that prevail in rice, but this chapter
focuses on leaf blast disease in rice crops. Blast disease is also known as
rice fever. Agricultural scientists find out about the disease by checking
their symptoms, similarly, this paper uses three particular symptoms for
leaf blast [5]. These symptoms will be used to form a decision tree and
49
50 Mathematics and Computer Science Volume 2
detect the labels of disease occurrence (as shown in Figure 4.1(a) and
Figure 4.1(b)).
The decision tree uses attributes to classify the leaves according to
symptoms. Every decision tree comprises of two types of nodes: a Decision
Node and the Decision Leaf Node. Different classification algorithms are
used nowadays to implement decision trees. Here, an ID3 classification
algorithm is used for implementation.
This algorithm stands for Iterative Dichotomiser 3. This algorithm forms
the smallest decision tree possible. The ID3 algorithm uses the Entropy and
Information Gain [6–8, 13]. The role of Entropy is to control the splitting
of the data for the decision tree. With the help of entropy, the boundaries of
the decision tree are decided. The formula for class entropy is given below:
−P P N N
T= log 2 − log 2 (4.1)
P+N P+N P+N P+N
−Pi Pi Ni Ni
IG (Pi , N i ) = log 2 − P + N log 2 P + N (4.2)
Pi + N i Pi + N i i i i i
E(S) =
∑ (P + N ) × I (P , N )
i i
G i i (4.3)
P+N
parameter value which has the highest information gain. The three symp-
toms taken for disease detection are the growth stage of the rice crop, disease
index, and lesion type [1]. Linguistic variables are taken for different ranges
of these variables. The ID3 algorithm will detect the leaf blast disease in rice
crops by using these symptoms as the attributes and decide from which
class the rice leaf belongs. The implementation of a decision tree in different
research is very vast. It framed different rules to form a decision tree that is
influenced by climatic parameters for the productivity of soybeans [2]. The
tree takes symptoms and forms forty rules for fuzzy logic for detecting leaf
blast disease by using regression methodology [3]. A decision tree classifier
is used by measuring the temperature and soil moisture as parameters to
form a system for predicting cotton crop disease [4]. Medical data mining
for the prediction of heart disease is done by using Naïve Bayes and decision
tree algorithms [8, 11, 12] and this research work shows a prediction accu-
racy of 99%. A prototype was used to evaluate the severity of disease on rice
crops by using computational intelligence and machine learning. K-Mean
segmentation has been used with fuzzy logic to calculate the degree of the
disease that occurred in rice plants [9, 10]. The prototype has 86% accuracy.
The attribute which has maximum gain value becomes the root node
and again the iteration of the whole process starts for expanding the tree
[21–23]. So, to form the table we have provided linguistic terms for differ-
ent values of each attribute or symptom. The rice plant has three important
growth stages and during these stages, the plant has some height between
52 Mathematics and Computer Science Volume 2
For the disease index, the range of input scores is varying from 0 to
100%. Scores are predefined in the standard evaluation system by IRRI
2015. The linguistic variables for different scales are provided below:
Lesion type (standard evaluation system of rice (SES), IRRI 2015) values
are taken from scale 5 to scale 9 because as per the experts, the lesions are
noticeable properly.
The result column has two specific classes for which two values are used
to provide disease labels:
Resistive (+) = 4
Susceptive (-) = 10
Detection and Classification of Leaf Blast Disease 53
This value is used to find the Gain for each attribute separately, as shown
in Table 4.2.
The choice of parameter is done to reflect the highest gain value for the
decision tree. Therefore, the ID3 algorithm is applied for creating a deci-
sion tree and we detect the symptoms which have maximum impact and
that will become the root of the tree, i.e., lesions type as it has the highest
information gain value. The attribute which has the greatest gain as the
decision node becomes the root of the decision tree so the tree would look
like a three-child leaf (Low, Mid, and High) with root node Lesions Type.
From Table 4.3, we get all mid and high-range lesion types having dis-
ease occurrence response as the susceptive (+) class label response, mean-
ing the farmers have to take measures to stop the severity of the disease.
But the low range of lesions type has two responses, both resistive and sus-
ceptive labels, for some rows so we take the next attribute branch which
has the second greatest information gain, i.e., disease index which is 0.46,
as depicted in Table 4.4.
Lesion Type
Lesion Type
Resistive Susceptive
Figure 4.2 has been derived by using two major symptoms of leaf blast
disease in which lesion type is the root node.
Now, in the next iteration we take the third greatest information gain
attribute branch, i.e., the growth stage with a gain value of 0.367. The tree
would take the shape, as shown in Figure 4.3.
4.4 Conclusion
Disease in crops and plants is a major problem faced by farmers. The tra-
ditional methods to detect the occurrence of diseases are tiresome, costly,
and are not time efficient in comparison with the new soft computing tech-
niques which provide a digital era in the agricultural sector. The decision
tree provides an easier view of the whole process and how to detect the
occurrence level of disease. We can say that the crop has specific symptoms
during a particular type of disease. This means by making use of symp-
toms, the risk level could be detected which is very helpful for farmers and
agricultural scientists to save the yield loss.
References
1. Pari Skamnioti, Sarah J. Gurr, Against the grain: safeguarding rice from rice
blast disease, Trends in Biotechnology, Volume 27, Issue 3, 2009, Pages 141-
150, ISSN 0167-7799, https://doi.org/10.1016/j.tibtech.2008.12.002.
2. P. Udupa, S. Vishwakarma, A survey of MRI segmentation Techniques for
Brain Tumor Studies, Bonfring International Journal of Advances in Image
Processing, Vol. 6, No. 3, August 2016, pp. 22-27. Doi: http://www.journal.
bonfring.org/papers/aip/volume6/BIJ-10467.pdf
3. F. M. Shamim, S. Vishwakarma, Exploiting the Motion Learning Paradigm
for Recognizing Human Actions, Bonfring Internation Journal of Advances
in Image Processing, Vol. 6, No. 3, August 2016, pp. 11-16. http://www.journal.
bonfring.org/papers/aip/volume6/BIJ-10465.pdf
4. Chopda, J., Raveshiya, H., Nakum, S., Nakrani, V., 2018. Cotton crop disease
detection using decision tree classifier. International Conference on Smart
City and Emerging Technology (ICSCET), 1–6.
5. Donatelli, M., Magarey, R.D., Bregaglio, S., L., W., Whish, J.P.M., Savary, S.,
2017. Modelling the impacts of pests and diseases on agricultural systems.
Agricultural Systems 155, 213–224.
6. Edwards-Murphy, F., Magno, M., Whelan, P.M., O’Halloran, J., Popovici,
E.M., 2016. b+wsn: Smart beehive with preliminary decision tree analysis for
agriculture and honey bee health monitoring. Computers and Electronics in
Agriculture 124, 211–219.
7. Ilic, M., Ilic, S., Jovic, S., Panic, S., 2018. Early cherry fruit pathogen disease
detection based on data mining prediction. Computers and Electronics in
Agriculture (Elsevier) 150, 418–425.
8. Karthiga, A.S., Mary, M.S., Yogasini, M., 2011. Early prediction of heart
disease using decision tree algorithm. International Journal of Advanced
Research in Basic Engineering Sciences and Technology (IJARBEST) 3(03).
9. Nithya, A., Sundaram, V., 2011. Wheat disease identification using classifica-
tion rules. International Journal of Scientific & Engineering Research 2(09),
1–5.
10. Sethy, P.K., Negi, B.S., 2018. Measurement of disease severity of rice crop
using machine learning and computational intelligence. Cognitive Science
and Artificial Intelligence (Springer) 2, 1–11.
11. Shouman, M., Turner, T., Stocker, R., 2011. Using decision tree for diag-
nosing heart disease patients. 9th Australasian Data Mining Conference
(AusDM 2011) 121, 23–29.
12. Singh, P., Jagyasi, B., Rai, N., Gharge, S., 1–6. Decision tree-based mobile
crowdsourcing for agriculture advisory system. 2014 Annual IEEE India
Conference (INDICON), 2014.
13. Zhang, Y.D., Wang, S., Dong, Z., 2014. Classification of alzheimer’s disease
based on structural magnetic resonance imaging by kernel support vector
machine decision tree. Progress In Electromagnetics Research 144, 171–184.
Detection and Classification of Leaf Blast Disease 57
Abstract
Emotion is significant in deciding a human’s ideas, behavior, and feelings. Using
the advantages of deep learning, an emotion detection system can be constructed
and various applications such as face unlocking, feedback analysis, and so on are
executed with high accuracy. Artificial intelligence’s fast development has made a
significant contribution to the technological world. However, it has several diffi-
culties in achieving optimal recognition. Interpersonal differences, the intricacy
of facial emotions, posture, and lighting, among other factors, provide signifi-
cant obstacles. To resolve these issues, a novel Hybrid Deep Convolutional based
Golden Eagle Network (HDC-GEN) model algorithm is proposed for the effective
recognition of human emotions. The main goal of this research is to create hybrid
optimal strategies that classify five diverse human facial reactions. The feature
extraction of this research is carried out using Heap Coupled Bat Optimization
(HBO) method. The execution of this research is performed by MATLAB soft-
ware. The simulation outcomes are compared with the conventional methods in
terms of accuracy, recall, precision, and F-measure and the comparison shows the
effective performance of proposed approaches in facial emotion recognition.
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (59–76) © 2023 Scrivener Publishing LLC
59
60 Mathematics and Computer Science Volume 2
5.1 Introduction
Emotional analytics combine psychology and technology in an intriguing
way. The examination of facial gestures is one of the most used methods
for recognition of emotions. Basic emotions are thought to be physiolog-
ically fixed, intrinsic, and universal to all individuals and many animals
[1]. Severe reactions are either a collection of fundamental emotions or a
collection of unusual feelings. The major issue is figuring out which emo-
tions are fundamental and which are complicated. While analyzing the
information acquired by the ears and sight, people may detect these signals
even when they are softly expressed. Based on psychological research that
illustrates that visual data affects speech intelligibility, it is reasonable to
infer that human emotion interpretation shares a predictable pattern [2].
The practice of recognizing human emotions from facial gestures is known
as recognition of facial expression. Facial expression is a universal signal
that all humans use to communicate their mood. In this era, facial expres-
sion detection systems are very important because they can record people’s
behavior, sentiments, and intentions [3].
Computer networks, software, and networking are rapidly evolving and
becoming more widely used. These systems play a vital part in our daily lives
and make life much easier for us. As a result, face expression detection as a
method of image analysis is quickly expanding. Human–computer interface,
psychological assessments, driving enabled businesses, automation, drunk
driver identification, healthcare, and, most importantly, lie detection are
all conceivable uses [4]. The human brain perceives emotions instinctively
and technology that can recognize emotions has recently been developed.
Furthermore, Artificial Intelligence (AI) can identify emotions by under-
standing the meaning of each face expression and learn to adapt to fresh input.
Artificial Neural Networks (ANN) are being employed in AI systems
currently [5]. Long Short Term Memory (LSTM) based RNN [6], Nave
Bayes, K-Nearest Neighbors (KNN), and Convolutional Neural Networks
[7] are used to tackle a variety of challenges such as excessive makeup,
position, and expression [8]. The properties of the swarm intelligence opti-
mization method, which can execute parallel computation and sharing of
information, have caught our interest in order to make good use of excess
computational resources and enhance optimization effectiveness [9]. The
Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO)
algorithms, as well as their upgraded variants, constitute the swarm intelli-
gence algorithm [10]. The traditional approaches are slow and inaccurate,
but a face expression detection system based on deep learning has proven
Human Face Emotion Recognition 61
Human emotion
dataset
Input image
Pre-processing
stage
by median filter
Feature
extraction by
Resize image
HBO method Face detection // knowledge
based algorithm
Performance
HDC-GEN method Emotion
measures
Parameter tuning
where n is denoted as time step, f1n is the optimal fitness speed, and the
best point is considered as X*. The extraction condition for each feature is
evaluated by the fitness of heap using Equation 5.4:
^
Mni n
Mni fin (t) , f Mni f f in (t )
n
fi ( t 1) (5.4)
^
f in n
Mni fin (t ) , f Mni f fin (t )
Initialization: Initialize the featured data as (aj, bj, cj). The kernel func-
tion is executed to the input function using Equation 5.5:
+∞
z=
(t ) =
m(t ) ∗ n(t ) ∑ m(τ )n(t − δ )
δ = −∞
(5.5)
where m is the input image feature and n is the convolution kernel perfor-
mance, τ is denoted as time delay, and t is symbolized as time, respectively.
The tangent hyperplane is evaluated for the kernel exploration by Equation
5.6:
a
h1c1 + h2c2 + ..haca= n ⇒ ∑h c = n
i =1
i i (5.6)
where h1, h2,……ha are the normal vector and c1, c2,……ca are the ith node
decision vector.
Max-Pooling Function: Once the kernel function is applied, then the
max-pooling operation is executed. The invariant features are extracted
in the kernel layer using the max-pooling function, turning the extracted
features into various images. The max-pooling function is applied in all
portions of the featured images using Equation 5.7:
x zj ,a = max
= p 1 (x z −1, a
( j −1)×o + p ) (5.7)
x zj = ∑h
a
α ( x zj1−1,a ) + baz −1
z −1
k, j (5.8)
Human Face Emotion Recognition 67
The training may be characterized using a vector that starts at the cur-
rent position of the data and ends at the ideal weight points in the data
memory. The training vector for classification is computed using Equation
5.9:
Start
Extract the features from the image using Initialized the features to
HBO method HDC-GEN for classification
No
If criteria met? New emotion exploitation is
apply
Yes
Stop
t
e x za
= =
f (x ) arg max m y(h | x ) = arg max m N
(5.10)
∑e
p =1
xt z p
only based about what it had previously learned and the image’s features.
As a result, for each image, it generated a list of identified emotion proba-
bilities. The number of correct forecasts was calculated by comparing the
highest probability reaction for each image with the real feelings associated
with the visuals.
Tp
Precision = ∧ ∧
(5.11)
T p+ Fp
∧
F p.
Accuracy: The accuracy is another important parameter for the evalua-
tion measures in emotion classification. It characterizes the accurate per-
centage of classified emotion and when the accuracy reaches 100%, then it
is referred to as the classification of emotion at its finest and is estimated
using Equation 5.12:
∧ ∧
T p+Tn
Accuracy = ∧ ∧ ∧ ∧
(5.12)
T p +Tn + F p + Fn
Tp
Sensitivity = ∧ ∧
(5.13)
T p + Fn
70 Mathematics and Computer Science Volume 2
2T p
F-Measure = ∧ ∧ ∧ (5.14)
2T p + F p + F n
100
80
GPSO-CNN
Precision (%)
60 BLSTM-RNN
SSA-DCNN
Proposed
40
20
0
Angry Sadness Fear Surprise Happy
The accuracy value of the proposed approach for anger, sadness, fear,
surprise, and happiness is compared with the existing methods GPSO-
CNN [15], BLSTM-RNN [13], and SSA-DCNN and is illustrated in
Figure 5.4. From this, the happy emotion achieved the highest accuracy
value by the proposed approach (99.2%) when compared with the exist-
ing methods GPSO-CNN (69%), BLSTM-RNN (86%), and SSA-DCNN
(95.33%).
Moreover, the recall value of the proposed approach for anger, sadness,
fear, surprise, and happiness is compared with existing methods such as
GPSO-CNN [15], BLSTM-RNN [13], and SSA-DCNN and is illustrated
in Figure 5.5. From this, the happy emotion achieved the highest recall
value by the proposed method (97.6%) when compared with the exist-
ing methods GPSO-CNN (80%), BLSTM-RNN (88%), and SSA-DCNN
(96%).
The F-measure value for the five different emotions compared with the
existing methods such as GPSO-CNN [15], BLSTM-RNN [13], and SSA-
DCNN is illustrated in Figure 5.6. From this, the happy emotion achieved
the highest F-measure value by the proposed method (97.98%) when com-
pared with the existing methods GPSO-CNN (85%), BLSTM-RNN (88%),
and SSA-DCNN (95%).
Also, the estimation of processing time is the foremost measure in this
study because less processing time is more significant than high time. The
processing time for the five different emotions compared with the existing
100
80
Accuracy (%)
60
GPSO-CNN
40 BLSTM-RNN
SSA-DCNN
Proposed
20
0
Angry Sadness Fear Surprise Happy
100
80
60
Recall (%)
GPSO-CNN
40 BLSTM-RNN
SSA-DCNN
Proposed
20
0
Angry Sadness Fear Surprise Happy
Proposed
Happy
SSA-DCNN
BLSTM-RNN
Surprise GPSO-CNN
Fear
Sadness
Angry
0 20 40 60 80 100
F-measure (%)
60
Processing time (ms)
40
GPSO-CNN
BLSTM-RNN
SSA-DCNN
20
Proposed
0
Angry Sadness Fear Surprise Happy
The error value obtained from the proposed classification model for
the five different emotions is compared with the existing methods such as
GPSO-CNN [15], BLSTM-RNN, [13] and SSA-DCNN and illustrated in
Figure 5.8. From this, the happy emotion achieved less error than the pro-
posed method (0.01%) compared with the existing methods GPSO-CNN
(1%), BLSTM-RNN (0.2%), and SSA-DCNN (0.18%).
1.8 GPSO-CNN
BLSTM-RNN
1.6 SSA-DCNN
Proposed
1.4
1.2
Error rate (%)
1.0
0.8
0.6
0.4
0.2
0.0
Angry Sadness Fear Surprise Happy
5.7 Conclusion
Emotion recognition is a difficult problem that is rising in relevance due
to its numerous applications. Facial expressions can be useful in assessing
a human’s emotion or mental state. The proposed design solutions were
executed effectively and a computational simulation verified their validity.
The development of a hybrid model and a program for identifying emo-
tions by facial expression is the major result of this study. Each category
of emotions has its own set of recognition accuracy metrics. Images with
the emotion “happy” (99.2%) have the best recognition accuracy, while
images with the emotion “sadness” have the worst (97%). Also, the com-
parative analysis proves that the proposed method has achieved superior
performance in emotion recognition over the existing models in terms of
high accuracy, precision, F-measure, recall, and less processing time as well
as error rate. In the future, a new intelligent algorithm can develop with
proper validation of various human emotions.
References
1. Ghanem, Bilal, Paolo Rosso, and Francisco Rangel. “An emotional analysis of
false information in social media and news articles.” ACM Transactions on
Internet Technology (TOIT) 20.2 (2020): 1-18.
2. Dasgupta, Poorna Banerjee. “Detection and analysis of human emo-
tions through voice and speech pattern processing.” arXiv preprint
arXiv:1710.10198 (2017).
3. Chen, Caihua. “An analysis of Mandarin emotional tendency recognition
based on expression spatiotemporal feature recognition.” International
Journal of Biometrics 13.2-3 (2021): 211-228.
4. Sánchez-Gordón, Mary, and Ricardo Colomo-Palacios. “Taking the emo-
tional pulse of software engineering—A systematic literature review of
empirical studies.” Information and Software Technology 115 (2019): 23-43.
5. Hemanth, D. Jude, and J. Anitha. “Brain signal based human emo-
tion analysis by circular back propagation and Deep Kohonen Neural
Networks.” Computers & Electrical Engineering 68 (2018): 170-180.
6. Du, Guanglong, et al. “A convolution bidirectional long short-term mem-
ory neural network for driver emotion recognition.” IEEE Transactions on
Intelligent Transportation Systems (2020).
7. Ashwin, T. S., and Guddeti Ram Mohana Reddy. “Automatic detection of stu-
dents’ affective states in classroom environment using hybrid convolutional
neural networks.” Education and Information Technologies 25.2 (2020):
1387-1415.
Human Face Emotion Recognition 75
Abstract
The pandemic of COVID-19, since its beginning, has been impacting the lives of
people the world over. However, in India, the impact would have been catastrophic
had Ayush’s intervention not been there. This paper examines the initiatives and
interventions by the leadership of the Ministry of Ayush (MoAyush) to manage
the pandemic situation through adaptive leadership. Effective communication,
use of IT platform, HR training, engagement with stakeholders, and adaptation
of R&D activities are the factors that helped the MoAyush to establish itself as the
leader in the fight against the pandemic of COVID-19.
6.1 Introduction
The very first case of COVID-19 was detected in China on 12th December
2019. A comprehensive package of technical guidance was issued by the
World Health Organization (WHO) with advice to all countries about
detection, testing, and managing potential cases on 10th January 2020. The
Govt. of India immediately directed the states and UTs to take all nec-
essary action for health sector preparedness following WHO’s advisory.
77
78 Mathematics and Computer Science Volume 2
the health budget but the impact among stakeholders has been large (PIB
release 7th Feb 2021).
The present government is very supportive of Ayush systems. Various
new and innovative programmes have been launched since 2014.
International Day of Yoga (IDY), the establishment of IT infrastructure in
the form of AyushGrid, the launch of National Ayush Mission, and certi-
fication of yoga professionals and accreditation of Yoga Institutions were
some of the significant contributions of the Ministry apart from regular
Central Sector Schemes. The Ayush services are available in all major gov-
ernment hospitals and Ayush treatment is part of the insurance policy now.
Out of 125,000 Health and Wellness Centres (HWCs) the government is
planning to establish, 10% would be dedicated to Ayush systems.
beds attached to 750 Ayush Colleges and 86 clinical facilities and Research
Councils under the Ministry. All the hospitals, colleges, centres, and facil-
ities were shared.
the world about Yoga and the holistic approach to health and wellbeing it
takes.
Despite IDY observances, though overcast by restrictions due to the
pandemic for a second time, the Ministry learned from previous years’
experience. The Ministry followed a three-pronged strategy which
included taking a digital-first approach, activating Government of India
stakeholders and their networks which have a wide-ranging reach, and
collaborating with the private sector in a greater way to pave the road for
the adoption of Yoga at a larger scale and development of the sector in
the years to come.
In the digital-first approach, the Ministry leveraged its existing IT
platforms and social media including Facebook, Twitter, Instagram,
YouTube, the Ministry’s website, and the Yoga Portal; it also leveraged
new platforms such as Koo, Josh, and Bolo India to reach wider viewers
which consume diverse kinds of content and prefer local and regional
languages. The Ministry released additional digital assets apart from
improvements in the existing ones, namely the Namaste Yoga appli-
cation. Further, the Government’s MyGov platform was used for the
first time to launch citizen-centred awareness building and engaging
activities. People were encouraged to practice yoga based on Common
Yoga Protocol (CYP) virtually across the country. Various IT tools and
online resources were created during the pandemic. Yoga has been very
useful for the well-being of people during the pandemic. The Prime
Minister of India himself leads the nation for IDY celebration. He
launched the “mYoga” application, developed in house by the Morarji
Desai National Institute of Yoga in collaboration with the World Health
Organisation (WHO) on 7th IDY on 21st June 2021. The app can be used
by all to promote one world, one health. (PIB release 21st June 2021).
The mYoga application focuses on WHO’s theme of ‘Be He@lthy, Be
Mobile (BHBM)’, which is in line with the United Nations’ Sustainable
Development Goal of achieving Universal Health Coverage by 2030.
The app addresses diseases and risk factors with an opportunity to nav-
igate upon the m-Health programme for Yoga. It offers a collection of
learning modules and practice sessions in both audio and video formats,
which enable users to practice Yoga anytime from the comfort of their
homes. The mYoga app is available in the Google Play Store and Apple
App Store. As per the AppBrain ranking report, (https://www.appbrain.
com/), the WHO mYoga app is placed at a Google Play store rating of
4.75. It has over 50+ thousand downloads.
86 Mathematics and Computer Science Volume 2
The yoga portal of the ministry had a viewership of over 1.05 million.
The Yoga Certification Board (YCB) for the Yoga Appreciation Programme
and Yoga Volunteer Training (YVT) Course, reported an outreach of
about 1.214 million. The Prime Minister addressed the nation on IDY
through various DD channels, emphasising the role of maintaining health
and well-being in the lives of people, especially at a time when the world
was reeling from the COVID pandemic.
All the Central Ministries, State Governments, public sector under-
takings, autonomous bodies, various yoga organizations, Universities,
National Cadet Corps (NCC), National Social Service (NSS), Nehru
Yuva Kendra (NYKs), schools, hospitals, railways, postal services, and
GramPradhans were brought on board. To reach the nooks and corners
of the country, MoAyush also used services of various Common Service
Centres (CSC) and Community Radio Stations to promote IDY activities
and motivate people to practice yoga. The Ministry also had a partnership
with Nickelodeon and HealthifyMe to reach different segments of people.
Indian Missions Abroad were also equally active in the promotion and
observation of IDY. The coordinated efforts of MoAyush have made Yoga
a mass movement and around 156.86 million people participated during
IDY 2021 (Table 6.1).
6.13 Conclusion
The crisis of COVID-19 has impacted the whole country, but for the Ayush
sector, it brought tremendous opportunities. The ministry of Ayush pro-
vided timely intervention and has helped reposition Ayush therapy not
only at the national level but worldwide. It was made possible with IT
tools and the use of the digital platform. Effective and regular communi-
cation during the continuity of the crises, sharing of resources with other
stakeholders, use of IT platform regressively, training and preparedness of
existing manpower, and leveraging old wisdom were some of the practices
which helped MoAyush to establish itself as a true leader at the time of
crisis.
References
1. Abrams, E. M., Shaker, M., Oppenheimer, J., Davis, R. S., Bukstein, D. A.,
&Greenhawt, M. (2020). The Challenges and Opportunities for Shared
Decision Making Highlighted by COVID-19. In Journal of Allergy and
Clinical Immunology: In Practice. American Academy of Allergy, Asthma
and Immunology. https://doi.org/10.1016/j.jaip.2020.07.003.
2. Boin, A., Kuipers, S., & Overdijk, W. (2013). Leadership in times of crisis:
A framework for assessment. International Review of Public Administration,
18(1), 79–91. https://doi.org/10.1080/12294659.2013.10805241
3. Kaul, V., Shah, V. H., & El-Serag, H. (2020). Leadership During Crisis:
Lessons and Applications from the COVID-19 Pandemic. Gastroenterology.
https://doi.org/10.1053/j.gastro.2020.04.076
4. Kaul V, Shah VH, EI-Serag H (2020). Leadership During Crisis: Lessons and
Applications from the COVID-19 Pandemic, Gastroenterology.
5. Lee A. Wuhan novel coronavirus (COVID-19): why global control is chal-
lenging? Vol. 179, Public Health. Elsevier B.V.; 2020. p. A1-2
6. Ministry of AYUSH, Govt. of India 2020. D.O. No. S.16030/18/2019-NAM
dated 6th March 2020, Ministry of AYUSH, Govt. of India 2020. Ayurveda’s
immunity-boosting measures for self-care during COVID 19 crisis. https://
www.ayush.gov.in/docs/123.pdf
7. Ministry of AYUSH, Govt. of India 2020.Guidelines for AYUSH Practitioners
for COVID 19. https://www.ayush.gov.in/ayush-guidelines.html
8. Pinsdorf, M. K. (2004). All Crises are Global: Managing to Escape
Chaos. United States: Fordham University Press.
9. Srikanth N, Rana R, Singhal R, Jameela S, Singh R, Khanduri S, Tripathi
A, Goel S, Chhatre L, Chandra A, Rao B, Dhiman K (2021). Mobile App–
Reported Use of Traditional Medicine for Maintenance of Health in India
Application of IT in Adaptive Leadership of MoAyush 89
Abstract
Proteins are arranged in a linear sequence due to peptide bonds. In proteins, a
peptide bond combines the amino group of one protein with the carboxyl group
of another protein. Protein secondary structure formation results from their bio-
physical and biochemical properties, like natural languages which depend on their
grammatical rule. So, the proposed model predicts a secondary structure from
protein primary sequences using the encoder-decoder based machine translation
method. The proposed model uses an encoder-decoder model based on long short-
term memory network. The proposed work uses training and testing performed on
available public datasets, namely CullPDB and data1199. The proposed model has
better Q3 accuracy of 84.87% and 87.39% for CullPDB and data1199, respectively.
Further, the proposed work was evaluated by comparing their performance with
other methods which predict secondary structure only from a single sequence. The
Encoder-Decoder Model for predicting secondary structure from a single primary
sequence is performing better than other single sequence-based methods.
Keywords: Protein structure prediction, amino acids sequence, proteomics,
one hot encoding, encoder-decoder, long short-term memory
7.1 Introduction
Protein is important biomolecules which help in developing the cells of
all living organisms. Proteins mainly perform the role of transporters,
catalysts, receptors, and hormones within all living organisms [1]. The
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (91–100) © 2023 Scrivener Publishing LLC
91
92 Mathematics and Computer Science Volume 2
=
out statet ΘOt (7.7)
where Wxa, Wxi, Whf, and Who are weight matrices for input vectors xt, Uha,
Uhi, Uhf, and Uho, which are weight matrices for previous state Outt-1, ba,
bi, bf, and bo, which are the bias terms for each memory gate at, it, ft, ot. ʘ
signifies the Hadamard product (element-wise product) operation and σ is
the sigmoid function.
The encoder-decoder model for the primary protein sequence to the
secondary structure is depicted in the Figure 7.1. In the model, the encoder
and decoder both use LSTM having 256 units. The state vector represen-
tation obtained from the encoder layer works as the input to the decoder
y1 y2 y3 y4
Predicted
Secondary
Structure
Decorder
Encorder
h1 h2 h3 hfinal hinitial
One hot Encoding
0 0 0 0 0 1 0 0
0 0 1 0 0 0 0 1
Encoding
One hot
0 1 0 0 0 0 0 0
0 0 0 1 1 0 0 0
1 0 0 0 0 0 1 0
Supervised
x1 x2 x3 x4 Secondary y1 y2 y3 y4
Primary Sequences Structure
Figure 7.1 Encoder–decoder model for protein primary sequence to secondary structure.
96 Mathematics and Computer Science Volume 2
LSTM. The encoder LSTM encoded the primary sequences, while the
second decoder LSTM predicts the secondary structure sequences. The
encoder LSTM layer processes the individual amino acid characters in
a sequence and finally, produce a state vector. The decoder receives the
state vectors from the encoder. After processing the complete input pri-
mary sequence, the decoder layer uses the secondary structure character to
decode the state vectors into a secondary structure sequence. The decoder
layer maximizes the value of log-likelihood of predicted secondary struc-
ture sequence on the basis of encoder layer state vectors and past protein
secondary structure characters.
The encoder layer processes the protein primary sequences by utiliz-
ing individual amino acid characters and finally, produce the state vectors,
while the decoder layer uses the secondary structure sequence and the
encoder layer state vectors.
7.5 Conclusion
In the proposed work, primary protein sequences are translated to their sec-
ondary structure using the sequence-to-sequence model. Protein primary
sequences are represented as one-hot encoding and directly feed to the
LSTM based sequence-to-sequence model. The proposed model has com-
paratively better performance on two publicly available datasets, CullPDB
and data1199. Despite its simplicity, the LSTM based sequence-to-sequence
98 Mathematics and Computer Science Volume 2
model easily captures the complex relationship between amino acids and
their secondary structure.
References
1. Z. Wang, F. Zhao, J. Peng, J. Xu, Protein 8-class secondary structure predic-
tion using conditional neural fields., Proteomics. 11 (2011) 3786–92. https://
doi.org/10.1002/pmic.201100196.
2. Q. Jiang, X. Jin, S.-J. Lee, S. Yao, Protein secondary structure prediction:
A survey of the state of the art., J. Mol. Graph. Model. 76 (2017) 379–402.
https://doi.org/10.1016/j.jmgm.2017.07.015.
3. Y. Wang, H. Mao, Z. Yi, Protein secondary structure prediction by using deep
learning method, Knowledge-Based Syst. (2017). https://doi.org/10.1016/j.
knosys.2016.11.015.
4. H. Schütze, S. Schütze, H. Adel, E. Asgari, arXiv:1610.00479v3 [cs.CL] 1 May
2017, 2017.
5. E. Vylomova, T. Cohn, X. He, G. Haffari, Word Representation Models
for Morphologically Rich Languages in Neural Machine Translation,
in: Proc. First Work. Subword Character Lev. Model. NLP, Association
for Computational Linguistics (ACL), 2018: pp. 103–108. https://doi.
org/10.18653/v1/w17-4115.
6. A. Drozdetskiy, C. Cole, J. Procter, G.J. Barton, JPred4: A protein second-
ary structure prediction server, Nucleic Acids Res. 43 (2015) W389–W394.
https://doi.org/10.1093/nar/gkv332.
7. D. Ganguly, D. Roy, M. Mitra, G.J.F. Jones, Word Embedding based
Generalized Language Model for Information Retrieval, in: 2015. https://doi.
org/10.1145/2766462.2767780.
8. M. Ashburner, C.A. Ball, J.A. Blake, D. Botstein, H. Butler, J.M. Cherry, A.P.
Davis, K. Dolinski, S.S. Dwight, J.T. Eppig, M.A. Harris, D.P. Hill, L. Issel-
Tarver, A. Kasarskis, S. Lewis, J.C. Matese, J.E. Richardson, M. Ringwald, G.M.
Rubin, G. Sherlock, Gene ontology: Tool for the unification of biology, Nat.
Genet. (2000). https://doi.org/10.1038/75556.
9. J. Chen, G. Liu, V. Pantalone, Q. Zhong, Physicochemical properties of pro-
teins extracted from four new Tennessee soybean lines, J. Agric. Food Res. 2
(2020) 100022. https://doi.org/10.1016/j.jafr.2020.100022.
10. M.N. Faraggi, A. Arnau, V.M. Silkin, Role of band structure and local-field
effects in the low-energy collective electronic excitation spectra of 2H-NbSe
2, Phys. Rev. B - Condens. Matter Mater. Phys. 86 (2012) 035115. https://doi.
org/10.1103/PhysRevB.86.035115.
11. C. Mirabello, G. Pollastri, Porter, PaleAle 4.0: High-accuracy prediction of
protein secondary structure and relative solvent accessibility, Bioinformatics.
29 (2013) 2056–2058. https://doi.org/10.1093/bioinformatics/btt344.
Encoder-Decoder for Protein Secondary Structure Prediction 99
Abstract
The vaccines for preventing COVID-19 are being considered as the most effec-
tive way to reduce the high pandemic burden on the global health infrastructure.
However, public hesitancy towards vaccination is a crucial and pressing problem.
Our study has been designed to determine the parameters affecting the vaccina-
tion decisions of common individuals. Using the platforms of the compartmental
model and network simulation, we categorize people and observe their motivation
towards vaccinations in a mathematical social contagion process. We consider
peer influence as an important factor in this dynamic and study how individuals
influence their peers regarding vaccination decisions. The efficiency of the vacci-
nation process is estimated by the period of time required to vaccinate a substan-
tial fraction of the total population. We discovered the major barriers and drivers
of this dynamics and concluded that it is required to formulate specific strategies
targeting specifically the undecided and vaccine hesitant people.
8.1 Introduction
The novel coronavirus disease 2019 (COVID-19), caused by severe acute
respiratory syndrome Coronavirus 2 (SARS-CoV-2), was first reported in
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (101–114) © 2023 Scrivener Publishing LLC
101
102 Mathematics and Computer Science Volume 2
well as the positive attitude of people towards vaccination [17]. In our study,
we focused on identifying the effects of the factors that could increase the
vaccination coverage through numerical simulations on an artificial society.
Moreover, we could justify several results found by survey based studies
where this kind of question has already been explored using survey results
[4, 15]. Our results address the vaccination acceptance problem using a
comprehensive mathematical and computational framework that would
help us figure out correct strategies to encourage community for vaccine
uptake and to stop further spreading of this pandemic.
ΥUV -μU
µ α(H+V)I βH
I H U
- µl -μH λH
V
-μV
In this model, the key variable we have is the vaccinated population who
are influencing other populations to get vaccinated. This is also the target
population which has to be maximized over a certain period of time. Here,
this group of people is playing an effective role to control the epidemic and
spread the infodemic. Considering the total population as 1 (normalized
form),
I+H+U+V=1
Hence, all the rate equations of this model are compiled as follows:
dI
= µ − α ( H + V )I − µ I (8.1)
dt
dH
= α ( H + V )I − β H − λ H + γ UV − µ H (8.2)
dt
106 Mathematics and Computer Science Volume 2
dU
= β H − γ UV − µU (8.3)
dt
dV
= λ H − µV (8.4)
dt
10000 VACCINATED
HESITANT
8000 UNWILLING
6000
4000
2000
0
0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0
This means that we consider a timescale, ζV, within which 90% of maxi-
mum vaccination, Vmax has been achieved. This will give us an estimate of
the vaccination coverage speed.
10000
Total unwilling population
λ=0.2
3000 λ=0.5
λ=0.8
8000 2500
6000 2000
4000 1500
1000
2000 λ=0.2
λ=0.5 500
0 λ=0.8
0
0 5 10 15 20 25 30 35 40
0 5 10 15 20 25 30 35 40
time time
(a) (b)
24
Total hesistant population
4000 λ=0.2 22
3500 λ=0.5
λ=0.8 20
3000
2500 18
2000 16
1500 14
1000 12
500 10
0 8
0 5 10 15 20 25 30 35 40
time 0.2 0.3 0.4 0.5 0.6 0.7 0.8
(c) (d)
Figure 8.3 (a) Vaccination coverage with time for different values of vaccination rate λ.
(b) Time evolution of unwilling people in population with different values of vaccination
rate parameter. (c) Time evolution of hesitant people in population with different values
of vaccination rate parameter. (d) Vaccination coverage time (ζV) for different values of
vaccination rate.
108 Mathematics and Computer Science Volume 2
10000
2500
4000 β = 0.2
β = 0.5
β = 0.8 12.5
3000 12.0
11.5
2000 11.0
10.5
1000
10.0
0 9.5
0 5 10 15 20 25 30 35 40 9.0
time 0.2 0.3 0.4 0.5 0.6 0.7 0.8
(c) (d)
Figure 8.4 (a) Vaccination coverage with time for different values of negative rumor
influence rate β. (b) Time evolution of unwilling people in population with different
values of negative rumor influence rate parameter. (c) Time evolution of hesitant
people in population with different values of negative rumor influence rate parameter.
(d) Vaccination coverage time (ζV) for different values of negative rumor influence rate.
someone who is vaccinated is motivating others who are not sure about
vaccination. Here, the vaccinated people influence the hesitant people to
get vaccinated. Here we are considering two parameters α and γ that take
into account this phenomenon. The results related to this are shown in
Figure 8.5 and Figure 8.6.
So, we can draw the conclusion that if more people are well informed
about the success rate of vaccination, then that would speed up the
110 Mathematics and Computer Science Volume 2
Y=0.5
2500 Y=0.8 14
2000 13
1500
12
1000
11
500
0 10
0 5 10 15 20 25 30 35 40 0.2 0.3 0.4 0.5 0.6 0.7 0.8
time (d)
(c)
Figure 8.5 (a) Vaccination coverage with time for different values of positive peer
influence rate γ. (b) Time evolution of unwilling people in population with different
values of positive peer influence rate parameter. (c) Time evolution of hesitant people in
population with different values of positive peer influence rate parameter. (d) Vaccination
coverage time (ζV) for different values of positive peer influence rate.
10000
Total vaccination population
2000 α= 0.2
Total unwilling population
α= 0.5
8000 α= 0.8
1500
6000
1000
4000
2000 500
α= 0.2
α= 0.5
α= 0.8
0 0
0 5 10 15 20 25 30 35 40
time 0 5 10 15 20 25 30 35 40
time
(a) (b)
4000 α= 0.2 16
Total hesitant population
α= 0.5
α= 0.8 15
3000
14
2000 13
12
1000
11
0 10
0 5 10 15 20 25 30 35 40
time 0.2 0.3 0.4 0.5 0.6 0.7 0.8
(c) (d)
Figure 8.6 (a) Vaccination coverage with time for different values of positive peer
influence rate α. (b) Time evolution of unwilling people in population with different
values of positive peer influence rate α. (c) Time evolution of hesitant people in
population with different values of positive peer influence rate α. (d) Vaccination coverage
time (ζV) for different values of positive peer influence rate α.
Hesitancy, Awareness, and Vaccination 111
References
1. W. Bank. The economy in the time of covid-19, 2020.
2. S. Bhattacharyya, A. Vutha, and C. T. Bauch. The impact of rare but severe
vaccine adverse events on behaviour-disease dynamics: a network model.
Scientific reports, 9(1):1–13, 2019.
3. A. A. Dror, N. Eisenbach, S. Taiber, N. G. Morozov, M. Mizrachi, A. Zigron, S.
Srouji, and E. Sela. Vaccine hesitancy: the next challenge in the fight against
covid-19. European journal of epidemiology, 35(8):775–779, 2020.
4. E. Dubé, M. Vivion, and N. E. MacDonald. Vaccine hesitancy, vaccine refusal
and the anti-vaccine movement: influence, impact and implications. Expert
review of vaccines, 14(1):99–117, 2015.
5. J. Eskola, P. Duclos, M. Schuster, N. E. MacDonald, et al. How to deal with
vaccine hesitancy? Vaccine, 33(34):4215–4217, 2015.
6. K. B. Habersaat and C. Jackson. Understanding vaccine acceptance
and demand—and ways to increase them. Bundesgesundheitsblatt-
Gesundheitsforschung-Gesundheitsschutz, 63(1):32–39, 2020.
7. C. Jarrett, R. Wilson, M. O’Leary, E. Eckersberger, H. J. Larson, et al. Strategies
for addressing vaccine hesitancy–a systematic review. Vaccine, 33(34):4180–
4190, 2015.
8. K. Kalimeri, M. G. Beiró, A. Urbinati, A. Bonanomi, A. Rosina, and C.
Cattuto. Human values and attitudes towards vaccination in social media. In
Companion Proceedings of The 2019 World Wide Web Conference, pages
248–254, 2019.
9. D. Kumar, R. Chandra, M. Mathur, S. Samdariya, and N. Kapoor. Vaccine
hesitancy: understanding better to address better. Israel journal of health
policy research, 5(1):1–8, 2016.
10. H. J. Larson, C. Jarrett, W. S. Schulz, M. Chaudhuri, Y. Zhou, E. Dube, M.
Schuster, N. E. MacDonald, R. Wilson, et al. Measuring vaccine hesitancy:
the development of a survey tool. Vaccine, 33(34):4165–4175, 2015.
11. J. V. Lazarus, S. C. Ratzan, A. Palayew, L. O. Gostin, H. J. Larson, K. Rabin, S.
Kimball, and A. El-Mohandes. A global survey of potential acceptance of a
covid-19 vaccine. Nature medicine, 27(2):225–228, 2021.
12. J. Luyten, L. Bruyneel, and A. J. van Hoek. Assessing vaccine hesitancy in
the UK population using a generalized vaccine hesitancy survey instrument.
Vaccine, 37(18):2494–2501, 2019.
13. J. C. Miller and T. Ting. Eon (epidemics on networks): a fast, flexible python
package for simulation, analytic approximation, and analysis of epidemics on
networks. arXiv preprint arXiv:2001.02436, 2020.
14. A. Nishi, G. Dewey, A. Endo, S. Neman, S. K. Iwamoto, M. Y. Ni, Y. Tsugawa,
G. Iosifidis, J. D. Smith, and S. D. Young. Network interventions for man-
aging the covid-19 pandemic and sustaining economy. Proceedings of the
National Academy of Sciences, 117(48):30285–30294, 2020.
Hesitancy, Awareness, and Vaccination 113
Abstract
This work investigates the waves propagation in an isotropic saturated porous-
thermoelastic medium. The incidence of a seismic wave is considered at the free
surface of a thermoelastic semi-infinite space. The surface of semi-infinite space
is taken as isothermal. Reflection ratios of waves for the incidence of longitudi-
nal waves are derived for impedance boundary conditions. The graph depicts the
effect of porosity on reflection ratios versus incident angle and impedance param-
eter. Graph plotting is done with the help of the MATLAB programming language.
Results are presented for a particular model in this paper.
9.1 Introduction
Thermoelasticity is the study of the change in shape and dimension of a
solid object when the temperature of the object varies. With the theory of
elasticity, Madan et al. [1] and Kumari and Madan [2] explore the propa-
gation of waves. By extending the classical coupled theory of thermoelas-
ticity by introducing relaxation time, Lord and Shulman [3] and Green
and Lindsay [4] presented the generalized theory of thermoelasticity. A
number of researchers [5–11] have discussed the phenomena of waves
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (115–136) © 2023 Scrivener Publishing LLC
115
116 Mathematics and Computer Science Volume 2
reflecting on a free surface. Rani and Madan [12] investigated the effect of
initial stress and imperfect interface on waves.
In porous media, the thermo-poroelasticity theory describes the rela-
tionship between temperature and stress components. A thermoelastic
saturated porous material with solid-fluid pores is a thermally conducting
porous solid and these materials are generally found in reservoir rocks of
the earth’s crust. The phenomenon of co-existence of thermoelasticity and
porosity plays a crucial role in the non-destructive evaluation of composite
materials and structures. The propagation of thermoelastic waves in fluid-
saturated soil was studied by Haibing [13]. Kumar et al. [14] investigated the
Rayleigh waves propagation in thermoelastic semi-infinite space with double
porosity under isothermal and insulated boundary conditions in thermo-
elastic semi-infinite space with double porosity. Wilson et al. [15] discuss the
propagation of acoustic waves. In a thermo-poroelastic medium, Wang et al.
[16] explored the reflection phenomenon of inhomogeneous elastic waves.
The reflection phenomenon of seismic waves at insulated and free surfaces of
thermo-poroelastic material was explored by Zorammuana and Singh [17].
Impedance boundary-conditions are defined as a linear combination
of unknown parameters and corresponding derivatives that are predict-
able on the surface. Electromagnetism, acoustics, seismology, science, and
technology all benefit from these conditions. The propagation of waves for
impedance boundary-conditions was studied by Singh [18], Vinh and Hue
[19], Vinh and Xuan [20], and others.
To conclude the wave propagation effects, a quantitative analysis of
reflection at the surface of a thermo-poroelastic medium is required.
This work examines the waves propagation at the surface of a saturated
porous-thermoelastic isothermal medium with an impedance boundary
condition. The influence of solid porosity and the Biot parameter for the
coupling of fluid and solid phases on the reflection ratios versus impedance
parameter and incident angle of a longitudinal wave is studied.
where
τij: components of stress,
Pf: pressure of fluid,
Propagation of Seismic Waves 117
β: Biot parameter,
δij: Kronecker delta
where
λ, μ: lame’s constants,
eij: components of strain,
ϴ: temperature,
γs: coefficient of thermal stress corresponding to solid,
γf: coefficient of thermal stress corresponding to fluid,
N: elastic parameter for bulk-coupling of fluid,
vi = f (Ui − ui),
f: solid porosity,
Ui: component of displacement corresponding to fluid,
ui: component of displacement corresponding to solid,
1 ∂u ∂u
Also, the strain components are emn = m + n , m, n = 1, 2, 3.
2 ∂xn ∂xm
The equation of motion in an isotropic thermal-porous solid saturated
with a non-viscous fluid are given by [24]:
σ=
ij , j ρ ui + ρ f vi , (9.4)
(P=
f ), i ρ f ui + qvi , (9.5)
∂
Kθ,ii= 1 + t 0 ( ρ ceθ + γθ0 (u k , k + v k , k ) ) , (9.6)
∂t
where
ρ: porous-aggregate density,
ρf: pore-fluid density,
q: coupling parameter between porous-aggregate solid and
pore-fluid,
γ = γs + βγf
118 Mathematics and Computer Science Volume 2
∂
Kθ,ii= 1 + t 0 ( ρ ceθ + γθ0 (u k , k + v k , k ) ) (9.9)
∂t
∂ps ∂qs ∂p f ∂q f
u1 = − , u3 = + (9.10)
∂x ∂z ∂z ∂x
∂p f ∂q f ∂p f ∂q f
v1 = − , v3 = + (9.11)
∂x ∂z ∂z ∂x
qq f + ρ f qs =
0 (9.15)
ρ ce (θ + t 0θ) + γθ0 (∇2 p s + t 0∇2 ps + ∇2 p f + t 0∇2 p f ) − K ∇2θ = 0 (9.16)
∂2 ∂2
where ∇2 = 2 + 2 is the Laplace operator.
∂x ∂z
The coupling structure of potentials pf , ps with thermal wave can be eas-
ily observed from Equations (9.12), (9.14), and (9.16), while the coupling
structure of potentials qf and qs can be easily observed from Equations
(9.13) and (9.15).
The velocities of waves are [17]
1
H1 H 2 − H 3 3
6 2
1
2 3
2 H H12
V1 = H 2 − H1 + −
2
−
3 9 H13 3 9
+ 3 2
−
27
H 1 H H
1 2 H 3
− +
27 6 2
1
H1 H 2 − H 3 3
6 2
1
2 3
2 H
H 2 − H1 + − 1 , (9.17)
3 9 H1 3 3
+ 3 2
−
27
H 1 − H1 H 2 + H 3
27 6 2
120 Mathematics and Computer Science Volume 2
( )
2
H2 H1
−
H1
3 9
V2 = −
( )
1
H H
{( ) ( )}
1
3
3 2
2
2 3 2 3
H3 H2 H1 H1 H1 H2 H3 H1
1 2
− + − + − + −
6 2 3 9 27 6 2 27
1
{( ) ( )}
1
3
3 2
H H
2 3 2 3
1 2
H3 H2 H1 H1 H1 H2 H3 H1
6 −
2
+
3
−
9 27
−
6
+
2
−
27
−
2
{ }
1
1 2
1
{( ) ( )}
3
2
H2
−
H1
3 2
2
3
H H
2 3 3
H3 H2 H1 H1 H1 H2 H3 H
3 9
+
6
1 2
− + − + − + −
1
1
2 3 9 27 6 2 27
H H
{( ) ( )}
1
3
3 2
2 3 2 3
H3 H2 H1 H1 H1 H2 H3 H1
1 2
− + − + − + −
i
6 2 3 9 27 6 2 27
+
2
(9.18)
2
H2 H1
−
3 9 H1
V3 = 1
−
3
1
3
2 H H H3 H H1
2
3
H 3
H1H 2 H3 H
2
2 3
+ + −
1 2 2 1 1
− − − +
6 2 3 9 27 6 2 27
1
1
3
H H H3 H H1
2
3
H 3
H1H 2 H3 H
2
2 3
+ + −
1 2 2 1 1
− − − +
6 2 3 9 27 6 2 27
− −
2
Propagation of Seismic Waves 121
{ }
1 2
H2 H1
32 −
3 9
1
+
1
H H
3
H3 H H1
2
3
H 3
H1H 2 H3
2
2 3
H1
+ + i
1 2 2 1
− − − + −
27
6 2 3 9 6 2 27
1
1
H
3
H
3 2
H H H3 H1
2
H 3
H1H 2 H3
2 3
+ + −
1 2 2 1 1
− − − +
6 2 3 9 27 6 2 27
(9.19)
2
µ
Vs = , (9.20)
ρ 2f
ρ−
q
η1 η2 η3 i
=
where, H1 = , H2 ,=
H3 = , t t0 + ,
η0 η0 η0 ω
x axis
Isotropic porous e1
thermoelastic semi- e e2
D1
infinite space (M) e3
es C3
C0 C2
C1
z axis
Figure 9.1 Geometry of problem for reflection of waves when longitudinal wave is
incident.
=ps C0e i (k0 ( xsin(e ) + zcos (e )) −ωt ) + C1e i (k1 ( xsin(e1 ) − zcos(e1 )) −ωt )
+C2e i (k2 ( xsin(e2 ) − zcos (e2 )) −ωt ) + C3e i (k3 ( xsin(e3 ) − zcos(e3 )) −ωt ) , (9.22)
=p f a0C0e i (k0 ( xsin(e ) + zcos (e )) −ωt ) + a1C1e i (k1 ( xsin(e1 ) − zcos(e1 )) −ωt )
+a C e 2 (
i ( k xsin ( e2 ) − zcos ( e2 ) ) −ωt )
2 2 + a C e i (k3 ( xsin(e3 ) − zcos (e3 )) −ωt ) (9.23)
3 3
b0C0e i (k0 ( xsin(e ) + zcos (e )) −ωt ) + b1C1e i (k1 ( xsin(e1 ) + zcos (e1 )) −ωt )
θ − θ0 =
+b2C2e i (k2 ( xsin(e2 ) − zcos (e2 )) −ωt ) + b3C3e i (k3 )( xsin(e3 ) − zcos (e3 )) −ωt ) , (9.26)
Propagation of Seismic Waves 123
where
( D − ρV02 ) γγ f − ( β N − ρ f V02 )
a0 = ,
γf
( N − qV02 ) − ( β N − ρ f V02 ) γ
b =
0
(( N − qV )( D − ρV ) − ( β N − ρ V ) ) k
0
2
0
2
f 0
2 2 2
0
,
2 γ f 2
( β N − ρ f V0 ) γ − ( N − qV0 ) γ
( D − ρVi2 ) γγ f − ( β N − ρ f Vi2 )
ai = , (9.27)
( N − qVi ) − ( β N − ρ f Vi ) γγ f
2 2
b =
i
(( N − qV )( D − ρV ) − ( β N − ρ V ) ) k
i
2
i
2
f i
2 2
i
2
, i = 1,2,3,
2 γ f
( β N − ρ f Vi ) γ − ( N − qVi2 ) γ
µ − ρVs2 2
d1 = 2 , D = λ + 2µ + β N
ρ f Vs
Using Snell’s Law, angles and corresponding wave vectors are associated
by:
k1 sin (e) = k1 sin (e1) = k2 sin (e2) = k3 sin (e3) = ks sin (es) (9.28)
2 2 γ b1 k12
( λ + β N ) + 2 µ cos e + a β N +
k12 k02
1 1
γ b k2
X1 + (λ + β 2 N ) + 2 µ cos 2 e2 + a2 β N + 22 22
k2 k0
γ b k2 k2
X 2 + (λ + β 2 N ) + 2 µ cos 2 e3 + a3 β N + 23 32 X 3 − 2 µ sin es cos es s2
k3 k0 k0
γb
X 4 = (λ + β 2 N ) + 2 µ cos 2 e + a0 β N + 20
k0
(9.30)
2k12 k1 2k22 k
2 sin e1 cos e1 − iω Z 2
sin e X +
1 1 2
sin e2 cos e2 − iω Z 22 sin e2
k0 k0 k0 k0
2k 2 k
X2 + 23 sin e3 cos e3 − iω Z 32 sin e3
k0 k0
k 2
k
X3 + (cos2 es − sin2 es ) s2 − iω Z 2s cos es
k0 k0
iω Z
X 4 = 2 sin e0 cos e0 + sin e0
k0
(9.31)
3. v 3 = 0 gives:
k1 k k k
a1 cos e1 X1 + 2 a2 cos e2 X2 + 3 a3 cos e3 X3 − s d sin es
k0 k0 k0 k0
(9.32)
X 4 = a0 cos e,
Propagation of Seismic Waves 125
∂θ
4. + hθ =0 (h → 0 for thermally-insulated surface, h → ∞
∂y
for isothermal surface) gives
where
C1 C2 C3 D1
=X1 = , X2 = , X3 = , X4 , (9.34)
C0 C0 C0 C0
BX = Y, (9.35)
where
2k 2 k
b23 = 23 sin e3 cos e3 − iω Z 32 sin e3 ,
k0 k0
k2 k
b24 = (cos 2 es − sin 2 es ) s2 − iω Z s2 cos es ,
k0 k0
k1 k k
b31 = a1 cos e1 , b32 = 2 a2 cos e2 , b33 = 3 a3 cos e3 ,
k0 k0 k0
k
b34 = − s d sin es , b41 = ( −ik1a1 cos(e1 ) + hb1 ),
k0
b42 = ( −ik2a2 cos(e2 ) + hb2 ), b43 = ( −ik3a3 cos(e3 ) + hb3 ),
γb
b44 = 0, z1 = (λ + β 2 N ) + 2µ cos 2 e + a0 β N + 20 ,
k0
iω Z
z 2 = 2sin e0 cos e0 + sin e0 , z3 = a0 cos e ,
k0
z 4 = −b0h + ik0 cos e.
(9.36)
PrefL1 R1 2 PrefL 2 R2 2
E1 = = X1 , E 2 = = X2 ,
Pinc L1 R0 Pinc L 2 R0
PrefL 3 R3 2 PrefT R4 2
E3 = = X 3 , E4 = = X4 , (9.38)
Pinc L 3 R0 Pinc T R0
Propagation of Seismic Waves 127
where PrefL1, PrefL2, PrefL3, and PrefT are the average energy of reflected coupled
longitudinal waves at e1, e2, e3, and transverse wave at e4, respectively, and
γ b γ f b0a0 β 3
R0 = − λ + 2 µ + β 2a0 N + a02 β N + s 2 0 + k0 cos e ,
k0 k02
γ b γ f b1a1 β 3
R1 = λ + 2 µ + β 2a1 N + a12 β N + s 2 1 + k1 cos e1 ,
k1 k12
γ b γ f b2a2 β 3
R2 = λ + 2 µ + β 2a2 N + a22 β N + s 2 2 + k2 cos e2 ,
k2 k22
γ b γ f b3a3 β 3
R3 = λ + 2 µ + β 2a3 N + a32 β N + s 2 3 + k3 cos e3 ,
k3 k32
R4 = µ ks3 cos es . (9.39)
J W
λ = 3.7 GPa, µ = 7.9 GPa, N = 6GPa, Ce = 1040 , K = 170 ,
kg K m
kg kg ρf
ρ = 2216 3 , ρ f = 950 3 , q = 1.05 ,
m m f
GPa
β f = 2.37 × 10 −3 , β s = 2 β f ,ω = 2s −1 ,
K
T0 = 300° K , t0 = 10 −10 s
The problem of isothermal surfaces has been studied here. To plot the
graphs and execute numerical calculations, the MATLAB programming
language is employed. Figure 9.2 explored the impact of porous-parameter
f on the reflection ratio of a coupled longitudinal wave at an angle e1 to the
incident angle e of the longitudinal wave. It is concluded that as the value of
f increases, the reflection ratio of a longitudinal wave decreases for e < 70°
128 Mathematics and Computer Science Volume 2
1.05
0.95
0.9
0.85
X1
0.8
0.75
0.7
0.65
0.6
0 10 20 30 40 50 60 70 80 90
e (in degrees)
Figure 9.2 Reflection ratio X1 of longitudinal wave (at angle e1) against e (incident angle)
for β = 0.04, Z = 10, f = 0.03 (circle-line), f = 0.04 (star-line), f = 0.05 (triangle-line).
and then for e > 70°, attaining the same value for f = 0.03 (circle-line), f =
0.04 (star-line), and f = 0.05 (triangle-line).
Variation of the reflection ratios of coupled longitudinal waves at an
angle e2, e3 against the incident angle for β = 0.4, Z = 10, f = 0.03 (circle-line),
f = 0.04 (star-line), f = 0.05 (triangle-line) is shown in Figures 9.3 and 9.4,
respectively. It is concluded that the reflection ratio of a coupled-longitu-
dinal wave against the incident angle attains minimum value at e = 71°,
then e > 71° attains the same value for distinct values of f. Qualitatively, the
×10–8
9
5
X2
4
0
0 10 20 30 40 50 60 70 80 90
e (in degrees)
Figure 9.3 Reflection ratio X2 of longitudinal wave (at angle e2) against incident angle e
for β = 0.04, Z = 10, f = 0.03 (circle-line), f = 0.04 (star-line), f = 0.05 (triangle-line).
Propagation of Seismic Waves 129
0.045
0.04
0.035
0.03
0.025
X3
0.02
0.015
0.01
0.005
0
0 10 20 30 40 50 60 70 80 90
e (in degrees)
Figure 9.4 Reflection ratio X3 of longitudinal wave (at angle e3) against incident angle e
for β = 0.04, Z = 10, f = 0.03 (circle-line), f = 0.04 (star-line), f = 0.05 (triangle-line).
4
3.5
3
2.5
X4
2
1.5
1
0.5
0
0 10 20 30 40 50 60 70 80 90
e (in degrees)
Figure 9.5 Reflection ratio X4 of transverse wave against incident angle e (in degrees) for
β = 0.04, Z = 10, f = 0.03 (circle-line), f = 0.04 (star-line), f = 0.05 (triangle-line).
130 Mathematics and Computer Science Volume 2
1.08
1.075
1.07
1.065
X1
1.06
1.055
1.05
1.045
1.04
0 5 10 15 20 25 30 35 40 45 50
Z
×10-8
6
4
X2
3
0
0 5 10 15 20 25 30 35 40 45 50
Z
0.014
0.013
0.012
0.011
X3
0.01
0.009
0.008
0.007
0 5 10 15 20 25 30 35 40 45 50
Z
0.5
0.45
0.4
0.35
X4
0.3
0.25
0.2
0.15
0.1
0 5 10 15 20 25 30 35 40 45 50
Z
Figure 9.9 Reflection ratio X4 of transverse wave against Z (impedance parameter) for
f = 0.01, e = 30°, β = 0.1 (circle-line), β = 0.2 (triangle-line), β = 0.3 (star-line).
1.1
0.9
0.8
0.7
E1
0.6
0.5
0.4
0.3
0 10 20 30 40 50 60 70 80 90
e (in degrees)
Figure 9.10 Energy ratio E1 of longitudinal wave at e1 against incident angle e for f = 0.05,
Z = 10, β = 0.3 (circle-line), β = 0.4 (star-line), β = 0.5 (triangle-line).
×10-8
4.5
4
3.5
3
2.5
E2
2
1.5
1
0.5
0
0 10 20 30 40 50 60 70 80 90
e (in degrees)
Figure 9.11 Energy ratio E2 of longitudinal wave at e2 against incident angle e for f = 0.05,
Z = 10, β = 0.3 (circle-line), β = 0.4 (star-line), β = 0.5 (triangle-line).
0.5
0.45
0.4
0.35
0.3
0.25
E3
0.2
0.15
0.1
0.05
0
0 10 20 30 40 50 60 70 80 90
e (in degrees)
Figure 9.12 Energy ratio E3 of longitudinal wave at e3 against incident angle e for f = 0.05,
Z = 10, β = 0.3 (circle-line), β = 0.4 (star-line), β = 0.5 (triangle-line).
Propagation of Seismic Waves 133
×10-3
3.5
2.5
2
E4
1.5
0.5
0
0 10 20 30 40 50 60 70 80 90
e (in degrees)
Figure 9.13 Energy ratio E4 of transverse wave against incident angle e for f = 0.05,
Z = 10, β = 0.3 (circle-line), β = 0.4 (star-line), β = 0.5 (triangle-line).
9.6 Conclusion
The effects of porosity and impedance on waves propagating in isotropic-
porous thermoelastic semi-infinite space is investigated in this work. The
following key points are drawn from graphs in this paper:
References
1. D.K. Madan, A. Rani and M. Punia, A note on the effect of rigidity and initial
stress on the propagation of Rayleigh waves in pre-stressed orthotropic elas-
tic layered medium, Proc. Indian Natl. Sci. Acad., Vol. 87, pp. 487–498, 2021.
2. A. Kumari and D.K. Madan, Deformation field due to seismic sources with
imperfect Interface, J. Earth Syst. Sci., Vol. 130, pp. 1-19, 2021.
3. H. Lord and Y. Shulman, A generalized dynamical theory of thermoelastic-
ity, J. Mech. Phys. Solids, Vol. 15, pp. 299-309, 1967.
4. A.E. Green and A. Lindsay, Thermoelasticity, J. Elast., Vol. 2, pp. 1-7, 1972.
5. B. Singh, Reflection of P and SV waves from free surface of an elastic solid
with generalized thermodiffusion, J. Earth Syst. Sci., Vol. 114, pp. 159-168,
2005.
6. A. N. Abd-Alla and A.A.S. Al-Dawy, The reflection phenomenon of SV waves
in a generalized thermoelastic medium, Int. J. Math. Sci., Vol. 23, pp. 529-
546, 2000.
7. J. N. Sharma, V. Kumar and D. Chand, Reflection of generalized thermo-
elastic waves from the boundary of a half-space, J. Thermal Stresses, Vol. 26,
pp. 925-942, 2003.
8. S. Abo-dahab, R.A. Mohamed and B. Singh, Rotation and magnetic field
effects on P wave reflection from a stress–free surface of elastic half-space
with voids under one thermal relaxation time, Journal of Vibration and
Control, Vol. 17, pp. 1827-1839, 2010.
9. R. Bijarnia and B. Singh, Propagation of plane waves in an anisotropic gen-
eralized thermoelastic solid with diffusion, Journal of Engineering Physics
and Thermophysics, Vol. 85, pp. 442-448, 2012.
10. S. B. Sinha and K. A. Elsibai, Reflection of thermoelastic waves at a solid half-
space with two thermal relaxation times, Journal of Thermal Stresses, Vol. 20,
pp. 129-146, 1996.
11. A.N. Sinha and S.B. Sinha, Reflection of thermoelastic waves at a solid
half-space with thermal relaxation, Journal of Physics of the Earth, Vol. 22,
pp. 237-244, 1974.
12. A. Rani and D.K. Madan, Effect of initial stress and imperfect interface on
love waves propagation in prestressed orthotropic layer coated over a pre-
stressed orthotropic semi-infinite space, Journal of Rajasthan Academy of
Physical Sciences, Vol. 20, pp. 219-228, 2021.
Propagation of Seismic Waves 135
Abstract
The objective was to make a comparative study of the IoT based Ensemble
Predictive system with real-life teacher predictions on student observation
during E-Learning. Data is collected from 46 faculties for 188 periods through an
opinion-based survey using a questionnaire. Similarly, for the 188 periods the data
was collected from the created IoT based Ensemble Predictive System. The system
is designed in such a way that it uses five variables, namely: Level of Interaction,
No. of Questions Raised, No. of Students in the Class, No. of Concepts Taught in
a Period, and Responsiveness of the Students to Questions during Class Hours, to
perform the student observation analysis. From observation, it is interpreted that
there is no major difference in the solution provided by the faculties and output
generated by the IoT Ensemble Predictive system. Further, it was found that there
is a remarkable relationship between the opinion provided by the faculties and
output generated by the IoT Ensemble Predictive system. But for the item (No.
of Students in the Class), there is no significant relationship between the opinion
provided by the faculties and output generated by the IoT Ensemble Predictive sys-
tem. Therefore, it can be interpreted that when the number of students increased
or decreased beyond ideal conditions, there is the possibility of deviation in the
opinion provided by the faculties and output generated by the system. Also, the
calculated R-value is 0.788, meaning there is a 78.8% strong positive relationship
between the opinion provided by the faculties and output generated by the system.
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (137–150) © 2023 Scrivener Publishing LLC
137
138 Mathematics and Computer Science Volume 2
10.1 Introduction
The World Wide Web was very efficient in terms of posting and retrieving
data in the beginning stages. The World Wide Web was only a collection of
static web pages with links to other sites and users could access the infor-
mation they needed by surfing and moving from one page to the next
[1]. After some time, the situation changed and web 2.0 was born. A user
of a website could do significantly more with it than they could with a
web 1.0 site, such as reading material and navigating through one page to
another page through hyperlinks. The normal person’s engagement with
online resources is easily available. Social networking sites like Facebook,
YouTube, and Flickr, among others, are examples of web 2.0 applications.
We are currently living in the age of semantic web. Computers in web 3.0
are capable of intelligently evaluating information available on web sites, as
well as developing and propagating the information they contain [2].
Until recently, the internet was designed to bring people together, enabling
them to communicate, exchange information, send messages, and hold vir-
tual conferences. We now have machine to machine interactions in addition
to human to human and human to machine interactions thanks to the intro-
duction of web 3.0. This is a huge step forward. System-to-system communi-
cation can be shown through a communication linking a temperature sensor
and AC [3]. Consider the following scenario: a temperature sensor monitors
the room temperature and when it rises beyond a certain threshold, it sends
an indication to the AC prompting it to immediately turn on the switch. If
the temperature goes down a specific point [4], the temperature sensor will
send a second signal to the air conditioner, this time instructing it to turn off
the machine. M2M is a technology that is used in the IOT. With the growth
of technology, it is now feasible for people to engage with and contribute
information on a website in a conventional manner. The information that
can be utilized in the reception and transmission of analyzed data may be
used to its full capacity by the specific kind of connection of different things
that can be achieved via the internet [5].
IoT Based Ensemble Predictive Techniques 139
10.3 Methodology
The descriptive and experimental research design is adopted for further
study. The data is collected from 46 faculties for 188 periods through an
opinion-based survey using a questionnaire. Similarly, the 188 periods the
data was collected from created an IoT based Ensemble Predictive System.
The system is designed in such a way that it uses five variables, namely:
Level of Interaction, No. of Questions Raised, No. of Students in the Class,
No. of Concepts Taught in a Period, and Responsiveness of the Students to
Questions during Class Hours, to perform the student observing analysis.
IoT Based Ensemble Predictive Techniques 143
From the Table 10.1 Group statistics - teacher & system feedback, mean
score calculated using the opinion provided by the faculties and output of
the analysis made by the system, it can be interpreted that students have
a high level of observation during the class hours, as the mean score lies
between (4.3404 - 4.4255).
Analysis is done to check the difference in opinion between the faculties
and system.
144 Mathematics and Computer Science Volume 2
From Table 10.2 Independent sample test - teacher & system feedback,
the approximated value is more than 0.05, meaning the null hypothesis is
accepted. Therefore, there is no major difference in the answers provided by
the faculties and output generated by the IoT Ensemble Predictive system.
Therefore, analysis is taken forward to identify whether there is any
significant relationship between the opinion provided by the faculties and
output generated by the IoT Ensemble Predictive system.
From Table 10.3 Correlation analysis - teacher & system feedback, the
approximate value is less than 0.05 for most of the items, meaning the null
hypothesis is rejected. Therefore, there is a closed relationship between
the opinion provided by the faculties and output generated by the IoT
Ensemble Predictive system.
But for the item (No. of Students in the Class), the evaluated value is
greater than 0.05 for most of the items, meaning the null hypothesis is
accepted. But, there is no serious relationship between the answers pro-
vided by the faculties and output generated by the IoT Ensemble Predictive
system when the No. of students increased or decreased beyond normal
ideal conditions.
Table 10.2 Independent sample test - teacher & system feedback. (Continued)
Independent samples test
Levene’s
test for
equality of
variances t-test for equality of means
F Sig. t df Sig. (2-tailed)
No. of students Equal variances .164 .686 -.977 374 .329
in the class assumed
References
1. Berecz, “Overview of E-learning Strategies from the Point of View of Higher
Education,” J. Appl. Multimed., vol. 13, no. 4, pp. 117–127, 2019, doi:
10.26648/jam.2018.4.003.
2. Hammi, R. Khatoun, S. Zeadally, A. Fayad, and L. Khoukhi, “IoT technolo-
gies for smart cities,” IET Networks, vol. 7, no. 1, pp. 1–13, 2018, doi: 10.1049/
iet-net.2017.0163.
IoT Based Ensemble Predictive Techniques 149
Abstract
This paper’s purpose is to model and analyze an impatient client renewal input
Bernoulli scheduled vacation interruption queue. Two types of congestion depen-
dent client impatience, namely balking and reneging, have been incorporated.
The server shifts to working vacation mode whenever the system is completely
depleted. The server follows Bernoulli’s vacation interruption schedule after exe-
cuting a service during working vacation, i.e., the server either continues working
during the vacation with certain probability or interrupts the working vacation
with the complementary probability. Regular service durations, working vacation
service durations, and vacation durations are expected to be congestion dependent
and exponentially distributed. The stationary probability distributions at different
epochs were estimated utilizing supplementary variable and iterative approaches.
A few performance metrics and the model parameters’ influence on the perfor-
mance metrics have been provided through numerical outcomes.
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (151–172) © 2023 Scrivener Publishing LLC
151
152 Mathematics and Computer Science Volume 2
11.1 Introduction
Queues or waiting lines are prominent in normal life and even in a range of
corporate and manufacturing contexts. Waiting for service is inescapable,
resulting in dissatisfaction, annoyance, and anxiety among clients. Clients
are strained for time in today’s fast-paced environment. They would pre-
fer not to be involved in the act of waiting for service. Clients’ behavior is
affected by such time limitations. Balking and reneging are two terms used
in queuing parlance to describe this type of impatience. If a client on arrival
discovers the queueing system is not empty, he or she may elect not to join it.
Balking is the term used in queueing theory to describe this behavior. Clients
who queue up but leave before the service is completed are termed to be
reneged. Clients’ balking behavior in an exponential queueing situation has
been initially proposed by Haight [1]. For the single server Markovian prob-
lem, Haight [2] introduced the reneging behavior of clients. The research of
a balking queue under renewal input process has been conducted by Finch
[3]. Wang et al. [4] carried out an optimum and sensitivity analysis as well as
a number of numerical experiments on a machine maintenance issue includ-
ing multiple servers and balking. Ward [5] examined the outcomes as well as
the reneging behavior of the queueing problem under traditional high traffic.
Queuing models with server vacations are especially significant for laying
the groundwork for effective design and research in a number of real life
scenarios like commercial and industrial manufacturing units, software and
information networking, and other sectors. Clients cannot access the service
facility for a predefined duration of time in vacation queueing approaches.
The server may be clearing surplus tasks left by departing clients, undergoing
restoration, or taking a break during vacation. Several authors have explored
queueing scenarios where the service facility takes a break, see [6−16].
Working vacation queues have gained popularity in the modern gener-
ations as a result of possible uses in data transmission, internet technology,
industrial systems, and many other associated systems. Servi and Finn [17]
presented the idea of working vacations (WV), wherein rather than shut-
ting down altogether, the service facility delivers service to the arrivals at a
decreased rate. This process is continued until a waiting client is discovered
at a vacation termination instant. A series of such WV is stated as multi-
ple working vacations (MWV). Working vacation queues are more conve-
nient to utilize since they are more accessible than the traditional vacation
queues. Due to this, working vacation queueing systems have sparked a
lot of interest. Banik et al. [18] published the analysis of an MWV queue,
wherein the arrivals follow a renewal process, utilizing a supplementary
Modelling and Analysis of a Congestion Dependent Queue 153
variable method. A finite buffer balking queue under renewal input process
has been reported by Vijaya Laxmi and Jyothsna [19]. Majid [20] worked
on the study of client impatience in a working vacation waiting line.
When some system indices reach a specific threshold in WV queues,
the server may restore regular service, interrupting the working vacation.
This sort of action is referred to as vacation interruption (VI) in working
vacation queues. Li and Tian [21] initially came up with the idea of VI
in a Markovian queue. Li et al. [22] extended Li and Tian [21]’s research
to a VI queue wherein the arrivals follow a renewal process. Majid and
Manoharan [23] published the research of a Markovian queue with VI
wherein the server is expected to interrupt the WV if the system detects at
least N waiting clients at a service finishing moment during the WV. At the
moment the service is completed in WV, the server determines whether to
extend the WV with a probability of q or to disrupt the WV with the com-
plimentary probability of 1 − q, referred to as Bernoulli scheduled vacation
interruption. Initially, Zhang and Shi [24] explored VI under the Bernoulli
rule in an exponential queue. The investigations of an impatient client VI
queue under Bernoulli rule and renewal input process has been published
by Vijaya Laxmi et al. [25]. More information regarding working vacation
queues with client impatience may be found in Bouchentouf and Yahiaoui
[26], Majid et al. [27], as well as references therein.
The majority of queueing system research is predicated on the notion
that the system’s state has no effect on the input and service parameters. In
practice, however, many of the queueing system’s service and input param-
eters are based on the system’s state and reducing the clients wait time.
Consider the case of telecommunications network routing. Congestion
independent routing probabilities cannot prevent tasks from being sent
to already overburdened disc stations in such networks, whereas jobs are
routed to the least loaded discs first when congestion dependent routing
probabilities are used. State dependent routing outperforms state inde-
pendent routing in terms of throughput. For literature on state dependent
queues, see [28−30] and the references therein.
To our knowledge, no research has ever examined queues that include
congestion dependence, working vacations, Bernoulli schedule VI, and
impatient clients; this gap in the literature drives the current research.
Also, the uncorrelated arrival process often yields a better approxima-
tion than the exponential distribution since the memoryless quality of the
aforementioned distribution might not necessarily fit the aims of applica-
tions and it can incorporate Erlang, deterministic distributions, etc. Hence,
investigating the impact of client impatience in a congestion dependent
Bernoulli scheduled VI queue under renewal input process is worthwhile.
154 Mathematics and Computer Science Volume 2
Service, as well as the vacation durations of the service facility, are expected
to be exponentially distributed and congestion dependent. The concept is
researched employing supplementary variable and iterative techniques.
The stationary distribution of the number of clients has been computed at
pre-arrival and arbitrary epochs. Performance metrics and numerical data
indicating the change in the performance metrics with the model parame-
ters are provided in table and graph forms.
The following is the layout of the article’s main body. Section 11.2 pres-
ents the model’s overview. The computation of the stationary distributions
at diverse time epochs is in Section 11.3 and special cases of the model
in Section 11.4. In Section 11.5, several model performance metrics are
provided and numerical calculations to show the parametric influence on
system performance metrics are disposed in Section 11.6. Finally, Section
11.7 brings the article to a conclusion.
∑ ∑ ∑
N N N
µj ηj φj
=j 1 =j 1 j =1
=µm = ; ηm ; φm =
N N N
At time t, let
π j (z )dz lim
= = Pr {N s (t ) j, z ≤ Z (t ) < z + dz}, z ≥ 0, 0 ≤ j ≤ N ,
t →∞
ω j (z )dz = lim Pr {N s (=
t ) j, z ≤ Z (t ) < z + dz }, z ≥ 0, 1 ≤ j ≤ N ,
t →∞
0
∞
=ω j (θ )
∫ e −θ zω j (z )dz 1 ≤ j ≤ N .
^
−π 0(1) (z ) =
µ1ω1(z ) + η1π 1(z ),
−π (j1) (z ) =
−r ( j)π j (z ) + a(z )((1 − β j )π j (0) + β j−1π j−1(0)) + (qη j+1
+ jα )π j+1(z ), 1 ≤ j ≤ N − 1,
−π N(1) (z ) =
−r (N )π N (z ) + a(z )(π N (0) + β N −1π N −1(0)),
−ω1(1) (z ) =
−s(1)ω1(z ) + a(z )(1 − β1 )ω1(0) + φ1π 1(z ) + s(2)ω2 (z )qη2π 2 (z ),
−ω (j1) (z ) =
−s( j)ω j (z ) + a(z )((1 − β j )ω j (0) + β j−1ω j−1(0)) + φ jπ j (z )
+ qηπ j+1(z ) + s( j + 1)ω j+1(z ), 2 ≤ j ≤ N − 1,
−ωN(1) (z ) =
−s(N )ωN (z ) + φN π N (z ) + a(z )(β N −1ωN −1(0) + ωN (0)),
where πj(0) and ωj(0) are respective rates of arrivals and for 1 ≤ j ≤ N, r(j) =
ϕj + ηj +(j − 1)α, s(j) = µj + (j − 1)α. The preceding set of equations is mul-
tiplied with e−θz and integrated over z for 0 to ∞ to get
−θ π 0 (θ ) =
−π 0 (0) + µ1 ω1(θ ) + η1 π 1(θ ),
^ ^ ^
(11.1)
^
−π j (0) + A(θ )((1 − β j )π j (0) + β j −1π j −1(0))
(r ( j) − θ )π j (θ ) =
^
+ (qη j +1 + jα )π j + 1(θ ),
^
1 ≤ j ≤ N – 1, (11.2)
^
(s(1) − θ ))ω1(θ ) =
−ω1(0) + A(θ )(1 − β1 )ω1(0) + φ1 π 1(θ )
^ ^
+ s(2)ω 2 (θ ) + qη2π 2 (θ ),
^
(11.4)
^
(s( j) − θ )ω j (θ ) =
−ω j (0) + A(θ )((1 − β j )ω j (0) + β j−1ω j−1(0))
^
+ φ j π j (θ ) + qηπ j+1(θ )
^
1 − Aˆ (θ )
∑ ∑ ∑ ∑
N N N N
πˆ j (θ ) + ωˆ j (θ ) = π j (0) + ω j (0 )
j =0 j =1 θ j =0 j =1
∑
N
Assigning θ → 0 and utilizing the normality condition πj +
j =0
∑
N
ω j = 1 produces
j =1
∑ ∑
N N
π j (0) + ωj =
λ (11.7)
=j 0=j 1
1 1
=π j− = π j (0); ω j− ω j (0). (11.8)
λ λ
^
1 − A(r (N ))
where ϱN = 0, ϱN−1 = ^ . Substituting πN−1 (0) in (11.3), we get
A(r (N ))
π N (θ ) = Ψ N ,θ π N (0),
^
158 Mathematics and Computer Science Volume 2
where
^
{A(θ )(β N −1N −1+N )−N
Ψ N ,θ = , if θ ≠ r (N ),
r (N ) − θ
^ (1)
− A (θ )(β N −1N −1 + N ), if θ =
r (N )
with
^ (l )
{A (θ )(β N −1N −1+N ) + lψ l −1
Ψ N ,θ = , if θ ≠ r (N ),
r (N ) − θ
^ (1)
− A (θ )(β N −1N −1 + N ), if θ =
r (N )
where
and
πˆ j (θ ) = Ψ j ,θ π N (0), j = N − 1, N − 2,…,1,
where
^
A(θ )(β j−1 j−1 + (1 − β j ) j ) + (qη j+1 + jα )Ψ j+1,θ −j
Ψ j ,θ = , if θ ≠ r ( j)
r (N ) − θ
^ (1)
− ( A (θ )(β j−1 j−1 + (1 − β j ) j ) + (qη j+1 + jα )Ψ (j1+)1,θ ), if θ = r ( j)
Modelling and Analysis of a Congestion Dependent Queue 159
with
where
^
1 − A(s(N )) φ Ψ
τ N 1,=
= ν N 0,τ N −1 = ^ , ν N −1 = − ^ N N ,s ( N )
A(s(N ))β N −1 A(s(N ))β N −1
ω N (θ ) ∆ N, θ ωN (0) + Ω N ,θ π N (0),
^
=
where
with
∧
(l )
(l ) A (θ )(β N −1τ N −1 + τ N ) − l∆(Nl −,θ1)
∆ N ,θ = if θ ≠ s(N ),
s( N ) − θ
∧ (l +1)
A (θ )(β N −1τ N −1 + τ N )
− , if θ = s(N ),
l +1
∧
(l )
(l ) A (θ )(β N −1ν N −1 + φN Ψ (Nl ),θ + lΩ(Nl −,θ1) )
Ω N ,θ = if θ ≠ s(N ),
s( N ) − θ
∧ (l +1)
A (θ )(β N −1ν N −1 + φN Ψ (Nl +,θ1) )
− , if θ = s(N )
l +1
where for 2 ≤ j ≤ N − 1,
and
ω̂
ω j (θ ) ∆ j, θ ω N (0) + Ω j ,θ π N (0), =
= j N − 1,…, 2,
where for 2 ≤ j ≤ N − 1,
∧ ∧
∧ ∧
with
∧
(l )
s( j + 1)∆(jl+)1,θ + A (θ )(β j −1τ j −1 + (1 − β j )τ j ) + l ∆(jl,θ−1)
∆ j ,θ = , if θ ≠ s( j),
s(l) − θ
∧
(l +1)
s( j + 1)∆(jl++11,θ) + A (θ )(β j −1τ j −1 + (1 − β j )τ j )
− , if θ = s( j),
l +1
(l ) ∧ (l ) (l ) (l )
(l ) s( j + 1)Ω j +1,θ + A (θ )( β j −1ν j −1 + (1 − β j )ν j ) + φ j Ψ j ,θ l Ω l j ,θ + q η j +1 Ψ j +1,θ
Ω j ,θ = , if θ ≠ s( j ),
s( j ) − θ
( l +1 ) ∧ ( l +1 ) ( l +1 ) ( l +1 )
s( j + 1)Ω j +1,θ + A (θ )( β j −1ν j −1 + (1 − β j )ν j ) + φΨ j ,θ + q η j +1 Ψ j +1,θ
− , if θ = s( j ),
l +1
ωN(0) = kπN(0),
where
∑ ∑
N N
π N (0) λ (
= j + (kτ j +ν j ))−1
j =0 j =1
162 Mathematics and Computer Science Volume 2
∑ ∑ (kτ +ν ))
N N
=π j− j ( j + j j
−1
,0 ≤ j ≤ N,
j =0 j =1
+ν )(∑ + ∑ (kτ
N N
−
ω
= j (kτ j j j j +ν j ))−1 ,1 ≤ j ≤ N .
j =0 j =1
Proof. Using (11.8) and (11.13) in (11.9) to (11.12), we obtain the result
of the theorem.
(λβ N −1 ) −
πN = π N −1 ,
r (N )
1
=πj ((qη j+1 + jα )π j+1 + λ (β j−1π j−−1 − β jπ j− )), =j N − 1,. . .,1,
r ( j)
1
=ωN (φN π N + λβ N −1ωN− −1 ),
s( N )
1
=ωj (s( j + 1)ω j+1 + φ jπ j + qη jπ j+1 + λ (β j−1π j−−1 − β jπ j− )),
s( j )
=j N − 1,. . ., 2
1
ω1 = (s(2)ω2 + φ1π 1 + qη2π 2 − λβ1π 1− )
s(1)
N
π0 = 1− ∑(π
j =1
j + ω j ).
Modelling and Analysis of a Congestion Dependent Queue 163
Ls = ∑ j(π + ω ); P = π
j =1
j j l
−
j + ω −j ; Ws = Ls /λe ,
N N
noticed in the table, which in turn results in least mean rate of client loss
(lr). This supports our decision to use βj = 1 − j/N 2 as the balking function.
Figure 11.1 and Figure 11.2 show the influence of λ on the expected
system length (Ls) and the mean rate of client loss (lr) for exponential inter-
arrival time distribution, respectively, for different q. The images show that
as λ is increased, both Ls and lr grow for any q. Furthermore, for a fixed λ,
2.6
2.4 q=0.0
q=0.4
2.2 q=0.8
q=1.0
2
1.8
Ls
1.6
1.4
1.2
0.8
0.5 1 1.5 2
λ
1.8
q=0.0
q=0.4
1.6
q=0.8
q=1.0
1.4
1.2
lr
0.8
0.6
0.4
1 1.5 2 2.5
λ
Ls and lr rise as q increases, meaning that Ls and lr are the least in VI queues
when compared to queues with no vacation interruption.
The impact of mean regular service rate (µm) on the expected system
length (Ls) and on the expected waiting duration (Ws) are displayed in
Figure 11.3 and Figure 11.4, respectively, for different mean vacation rates
(ϕm), wherein the inter-arrival durations are considered to be Erlang-5 dis-
tributed. As expected, Ls and Ws deplete with the increase of µm for any ϕm.
A similar effect is seen on Ls and Ws with the increase of ϕm for fixed µm.
Figures 11.5 and 11.6 show the changes in lr and Ws, respectively, while
the inter-arrival times are distributed according to deterministic distribu-
tion for different ηm and ϕm. It is worth noting that as ηm and ϕm grow, lr
and Ws deplete.
For different inter-arrival time distributions, Figure 11.7 displays the
system capacity (N) versus the blocking probability (Pl). The blocking
probability (Pl) is lowest when the inter-arrival durations are distributed
according to deterministic distribution and highest when they are distrib-
uted according to HE2 distribution.
When inter-arrival durations are exponentially distributed, the impact
of the reneging rate (α) on the expected system length (Ls) and on the mean
rate of client loss (lr) is depicted in Figure 11.8. As visualized in the graph,
the expected system length falls as α grows, but the mean rate of client loss
increases with α, as would be predicted.
2.25
φm = 0.55
2.2 φm = 0.82
φm = 1.10
2.15
2.1
2.05
Ls
1.95
1.9
1.85
1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7
µm
1.46
φm=0.55
1.44 φm=0.82
φm=1.10
1.42
1.4
1.38
1.36
Ws
1.34
1.32
1.3
1.28
1.26
1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7
µm
0.81
φm = 0.55
0.8
φm = 0.82
φm = 1.10
0.79
0.78
0.77
lr
0.76
0.75
0.74
0.73
0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6
ηm
1.4
φm = 0.55
1.39
φm = 0.82
φm = 1.10
1.38
1.37
1.36
Ws
1.35
1.34
1.33
1.32
0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6
ηm
0.12
HE2
Exponential
0.1
Erlang 3
Deterministic
0.08
Pl
0.06
0.04
0.02
0
4 5 6 7 8 9 10 11 12 13 14 15
N
3.5
Ls
3 Ir
2.5
1.5
0.5
0
0.5 1 1.5 2 2.5
α
11.7 Conclusion
The current research work examines a finite capacity renewal input con-
gestion dependent queue with Bernoulli scheduled vacation interruption
and client impatience. The regular service rates, working vacation service
rates, and vacation rates are considered to be exponentially distributed and
are congestion dependent. To obtain the stationary probabilities at various
epochs, we have adopted supplementary variable and iterative techniques.
We also presented a few numerical simulations to show the consequence of
the model parameters on the key metrics, which are provided in table and
graph forms. The current findings would be beneficial and significant for
modelling the transportation business, manufacturing systems, medical
administration, etc. The future direction of this work would be to expand
the model under examination by incorporating state dependent inter-
arrival times.
References
1. F. A. Haight, Queueing with balking (1957), Biometrika, vol.44, issue.3-4,
pp.360-369, https://doi.org/10.1093/biomet/44.3-4.360
170 Mathematics and Computer Science Volume 2
Abstract
Living cells need a constant availability of certain resources to have a sustained gene
expression process. Limited availability of cellular resources for gene expression,
like ribosomes and RNA Polymerase, significantly modifies the system dynamics.
Factors like the variation in rate of binding or variation in efficiency of the recruited
resource have the potential to affect crucial dynamic phenomena like cell fate deter-
mination. In this paper we have taken a very important motif, a bistable genetic tog-
gle switch, and explored the effect of resource imbalance in this circuit in terms of
the bifurcations taking place. We show that initial asymmetric biasing to resources
via resource affinity or gene copy number significantly modifies the cell fate tran-
sition, both in pitchfork and saddle node type bifurcation. Our study establishes
that in a limited resource environment, controlled resource allocation can be an
important factor for robust functioning of the synthetic or cellular genetic switches.
12.1 Introduction
Proteins that govern all functionalities in a living cell are produced by two
major steps: transcription and translation, which indeed are a combination
of several intermediate processes. In order to understand intriguing cellu-
lar processes like cellular decision making or to operate a synthetic circuit
inside the cell, a detailed mathematical study of cellular or synthetic gene
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (173–186) © 2023 Scrivener Publishing LLC
173
174 Mathematics and Computer Science Volume 2
resource the gene circuit collects from the cell. Though the presence of this
essential cellular resource in all the prokaryotes and eukaryotes is a fact,
the amount of free ribosome in different living beings is different. Even for
different cells having different functionalities, they differ in availability of
free ribosomes. Like the pancreatic cells, in eukaryotes, they are dedicated
for most of the protein production over other cells and contain an unusu-
ally high number of ribosomes. On the contrary, the smooth endoplasmic
reticulum (SER) does not have ribosomes on its surface and thus does not
participate in protein production. This limited availability of the essen-
tial translational resource inside the cell significantly affects the ongoing
dynamics. In a low protein activity or for a demand of low resource for the
implemented synthetic construct, this limited availability may not affect
things, but in higher protein activity or for a larger resource demand for
the synthetic construct, unprecedented resource competition comes into
the picture. Different experimental and theoretical studies establish that
ribosome limitation, significantly modifying the circuit dynamics and at
a larger scale the system chooses the favorable state. In a recent study, it is
established that the protein production curve can be largely modified in
terms of sensitivity and amplification by controlling ribosome availabil-
ity and its distribution in the system [5]. Due to the nature of coupling
of different ongoing processes, the competition in outer motifs, which
seems unimportant in the study of motif interest, significantly modifies the
dynamic behavior and can ruin the entire system, resulting in some emer-
gent responses in output [4]. Not only limited to ribosomes, this competi-
tion can arise for RNAP, gene copy numbers, degradation machinery, and
many more. Yuriy Mileiko and team in a recent study have shown that gene
copy number variation brings a significant change in the dynamics of some
well-known motifs [13]. The effect of decoy binding is also established by
some recent papers [9, 11]. This trend of exploring the consequences of
resource availability and distribution in cell dynamics is new and scientists
are getting exciting novel mechanisms of system modelling from it.
Cellular decision-making, environmental sensing, and cell to cell com-
munication are three key processes underlying pattern formation and
development in microscopic and complex organisms. From a theoretical
point of view, though the cellular decision making seems to be reversible,
in practice most of the time this is irreversible due to some secondary
effects arising from the process. The biologically programmed cell death
(apoptis) or cell death in response to injury (lysis) is most the promising
reason here. Several approaches in determining cell fate like feedback-con-
trolled regulation, cell size, and growth rate dependency [10, 17] are estab-
lished in the recent past. The effect of noise, especially when a group of cells
176 Mathematics and Computer Science Volume 2
X Y
resx resy
Resource T
(a) (b)
Steady states of protein
Steady states of protein
Bistable
region
Monostable Monostable
region region Monostable Bistable
region region
(c) (d)
Figure 12.1 Model Motif. (a) Genetic toggle, two proteins X and Y mutually repress each
others promoter activity. Hammerhead symbol represents the repression here. Also, both
the participants of the toggle collect resources from the same pool T with the affinities resx
and resy respectively. (b) Schematic diagram of inter cellular competition determining cell
fate. Strength of repression drives the cell to a particular fate. (c) Saddle node bifurcation
in genetic toggle. A bistable region separates two monostable regions in phase space.
(d) Pitchfork bifurcation in genetic toggle. Monostable region switches to a bistable region
at bifurcation point.
The available mRNA pool gx and gy, collecting ribosomes from the pool
T, with affinities resx and resy, makes a ribosome bound complex cx and cy
which will be translated to protein X and Y at a rate of ∈x and ∈y, respec-
tively. This asymmetry in resource allocation is very insightful here. The
polycistronic mRNA pool in most of the bacterial organisms contains mul-
tiple ribosome binding sites (RBS) [3] and the rate of translation depends
on the rate of recruitment of ribosomes to this RBS, as well as on the rate
of translation initiation. The rate of ribosome recruitment also depends
178 Mathematics and Computer Science Volume 2
upon many factors. Including all these rates, we generalize the resource
allocation or the protein production, which can have different rates as well.
As mentioned, from the total pool T, the ribosome bound complexes
are presented by cx and cy, thus further free pool of resource ribosome for
translation is estimated by (T − cx − cy).
The mutual repression is captured by a Michaelis–Menten type term in
our model and the hill function co-operativity n is taken as 2. The ODE
representing the scenario is given by Equation 12.1 below.
dc x
= res x (T − c x − c y ) g x − c x
dt
dX c x ∈x
= −X
dt 1 + Y n
(12.1)
dc y
= res y (T − c x − c y ) g y − c y
dt
dY c y ∈y
= −Y
dt 1 + X n
In a steady state, all the rates of change are equal to 0 and we investigated
the system.
7
10
Concentration of protein X
Concentration of protein X
4 6
3 4
2
2
1
1 2 3 4 5 1 2 3 4 5
Rate of X production (Єx) Rate of X production (Єx)
(a) (b)
12 14
rex = 2
Concentration of protein X
rex = 3
Concentration of protein X
10 12
8 10
8
6
6
4
4
2 2
1 2 3 4 5 1 2 3 4 5
Rate of X production (Єx) Rate of X production (Єx)
(c) (d)
6 8
Concentration of protein X
Concentration of protein X
gx = 3 gx = 4.5
5 6
4
3 4
2 2
1
1 2 3 4 1 2 3 4
Rate of X production (Єx) Rate of X production (Єx)
(a) (b)
10
Concentration of protein X
Concentration of protein X
8
gx = 5 8 gx = 6
6
6
4
4
2 2
1 2 3 4 1 2 3 4
Rate of X production (Єx) Rate of X production (Єx)
(c) (d)
Figure 12.3 Availability of Total mRNA Pool Regulates Symmetry of Pitchfork Bifurcation
and Cell Fate Transition. Concentration of protein X vs. rate of X production ∈x plot. resx =
resy = 2, n = 2, T = 5, gy = 5 is fixed for all the plots and ∈x = ∈y. Gene copy numbers available
for production of X, i.e., gx =3 for (a), gx = 4.5 for (b), gx = 5 for (c), and gx = 6 for (d).
7
Concentration of protein X
Concentration of protein X
T=5 14
6 T=10
12
5 10
4 8
3 6
2 4
1 2
0.5 1.0 1.5 2.0 2.5 3.0 0.5 1.0 1.5 2.0 2.5 3.0
Rate of X production (Єx) Rate of X production (Єx)
(a) (b)
1.6
10 1.5
Range of bistable region (Єx)
Concentration of protein X
1.4
8
1.3
6 resx = 1 1.2
resx = 2 1.1
4 resx = 3
1
0.9
2
0.8
0 0.7
0 2 4 6 8 5 5.5 6 6.5 7 7.5 8 8.5 9 5.5 10
Rate of X production (Єx) Available Total resource T
(a) (b)
Figure 12.5 (a). Rate of X production ∈x vs. concentration of protein X plot in case
of saddle node bifurcation in genetic toggle. resy = 2, gx = gy = 5, T = 5, n = 2, ∈y = 2.
(b). Phase plot for range of bistable region ∈x, the rate of X production from its complex
vs. total resource availability T. resx = resy = 1, n = 2, gx = gy = 5, ∈y = 1.
Alternate Cell Fate in Bistable Genetic Switch 183
12.4 Conclusion
Cellular decision making is a fundamental biological phenomenon by
which a cell opts the different states prior to environmental conditions,
leading to asymmetric cell differentiation. The underlying reason behind
this is still not entirely explored. We take a simple genetic motif, genetic
toggle here, and show that resource affinity asymmetry of the toggle par-
ticipants, both for the saddle node and in pitchfork bifurcation, signifi-
cantly biases the cell fate. This mutually repressing motif is very common
in nature [6, 14], where in output the system shows patterning by choosing
one cell fate over another. The availability of total resource pool signifi-
cantly regulates the bifurcation point in the motif. In the case of a synthetic
circuit, this resource limitation is very true because the gene circuit imple-
mented in the host entirely depends upon the host’s resource for its expres-
sion and in the case of a cell, the limitation in cellular resource ribosome
is a major factor, indicating our findings to be true also. We also investi-
gated the effect of gene copy number in the case of pitchfork bifurcation,
indicating an initial asymmetry that biases the cell fate to lower or higher
production states accordingly.
Here, it is important to note that our entire consideration is valid for a
low growth state of the system. The effect of growth rate on cell dynamics
is well established [17]. Overexpression of endogenous genes, or adding
184 Mathematics and Computer Science Volume 2
Acknowledgement
PC and SG acknowledge the support by DST-INSPIRE, India, vide sanc-
tion Letter No. DST/INSPIRE/04/2017/002765 dated- 13.03.2019.
References
1. G. Balázsi, A. van Oudenaarden, and J. J. Collins. Cellular decision mak-
ing and biological noise: from microbes to mammals. Cell, 144(6):910–925,
2011.
2. I. Bose and S. Ghosh. Bifurcation and criticality. Journal of Statistical
Mechanics: Theory and Experiment, 2019(4):043403, 2019.
3. D. H. Burkhardt, S. Rouskin, Y. Zhang, G.-W. Li, J. S. Weissman, and C. A.
Gross. Operon mrnas are organized into orf-centric structures that predict
translation efficiency. Elife, 6:e22037, 2017.
4. P. Chakraborty and S. Ghosh. Emergent correlations in gene expression
dynamics as foot-prints of resource competition. The European Physical
Journal E, 44(10):1–12, 2021.
5. P. Chakraborty and S. Ghosh. Emergent regulatory response and shift of half
induction point under resource competition in genetic circuits. arXiv pre-
print arXiv:2112.04985, 2021.
Alternate Cell Fate in Bistable Genetic Switch 185
Abstract
In last few decades, researchers have been motivated to facilitate ontology in sev-
eral fields. In medical science, ontology is used to describe the theory of medical
vocabularies and the correlation shared among them, thus permitting the sharing
of medical knowledge. Definition based on ontology of disease allows each class of
disease to be classified in a formalized structure singularly and along with the dis-
cussion of ontological realism for the treatment and diagnosis of disease. This paper
uses a hybrid approach to evaluate ontology for epilepsy disease. A multi-criteria
decision making (MCDM) method is used to decide the best ontology provided by
Bio-portal to select the best suitable characteristics for epilepsy disease.
13.1 Introduction
Nowadays, the single resource most critical for top management in an
organization to sustain competitive advantage is knowledge. An orga-
nization can have competitive edge over its competitors by building an
excellent process to manage knowledge. Ontology is a static knowledge
representation method that includes definitions of basic concepts in a
domain and relations among them which are machine interpretable. The
information retrieval quality is improved from a keyword-based retrieval
to knowledge-based search. Users should have a way for assessing ontolo-
gies and deciding which one fits their requirements to best face a multitude
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (187–204) © 2023 Scrivener Publishing LLC
187
188 Mathematics and Computer Science Volume 2
13.2 Background
Methods have been developed to evaluate and select the best ontology. [16]
explored the gain utilization of Semantic Web technologies in the domain
of recruitment and developed an ontology based recruitment process. [17]
used ontology evaluation techniques for agent cooperation to measure the
quality of ontology. A criteria selection framework was proposed by [10]
for guiding the selection of suitable criteria for various levels of ontology
evaluation. [5] presented a method for ontology evaluation based on the
goal, question, and metric approach for empirical evaluation. [15] pro-
posed a method to integrate ontologies to select and recommend adapted
internship seekers. [22] presented metrics, approaches, and other similar
aspects of ontology evaluation in a concise manner.
[12] presented a scalable data-driven framework for ontology evaluation,
targeting Big Data scenarios and use cases. [21] presented the outstanding
A Hybrid Approach to Ontology Evaluation 189
13.4.1 Accuracy
To obtain higher accuracy, correct definitions and descriptions of classes,
properties, and individuals which clearly define the domain are required,
i.e., the considered ontologies should specify epilepsy disease.
13.4.2 Adaptability
Adaptability defines how long the ontology predicts its uses. Ontologies for
epilepsy are originally designed to describe the criteria on epilepsy disease
and its vocabulary also allows formalizing symptoms of all kinds and dif-
ferentiating between diseases.
13.4.3 Clarity
Clarity calculates how productively the ontology can communicate with
the deliberate meaning of the given terms. The name of the ontology
should clearly explain the content and its function.
192 Mathematics and Computer Science Volume 2
13.4.4 Completeness
For completeness, the domain of interest and the thickness and rich-
ness of the ontology should be properly covered. In order to identify the
disorder, the list of all relevant characteristics is provided by the three
ontologies.
13.4.5 Conciseness
Conciseness is the evaluation criteria that states if the ontology includes
irrelevant elements regarding to the domain to be covered. For example,
ontology about epilepsy disease may take an important view on what the
disease actually is. It is not important to state if a person suffering from
other disease and any related information about that.
13.4.6 Consistency
Consistency explains that the ontology should not involve or allow for any
discrepancies, for example: Confused being Confused Memory is the one
of the symptoms of epilepsy, but having a logical axiom, for example: call-
ing confused a mental state will contradict the statement.
EPILEPSY
A1 A2 A3 A4 A5 B1 B2 B3 B4 B5 C1 C2 C3 C4 C5 D1 D2 D3 D4 D5 D6 D7
Continuant
Dependent Continuant
Diease
Epilepsy
Focal Epilepsy
Classification_System
Blume_2001_ILAE_Glossary
Blume_2001_I_General_Terms
Blume_2001_I_5.0_Focal
Focal
A Hybrid Approach to Ontology Evaluation 197
Crises Continuas
According to above data, the focal type of epilepsy exists in all the three
ontologies. The focal type of epilepsy is the most prominent type among
all the epilepsy types. Most patients suffering from epilepsy suffer from
this type of the disease. In EPSO type presents with 7-class mapping, ESSO
presents a focal class with 16-class mapping, and EPILONT presents with
0-mappping. Considering the above data, pairwise comparison matrixes
are generated and Ontology Weights (Ow) are computed using the Eigen-
vector approach.
The results are consistent, as the values of CI and CR were less than 0.1.
The same procedure is performed for all other selected criteria.
For Cluster II, Treatment, the class mapping of each of the considered
Epilepsy Ontology was observed. The Treatment AED is present in class
mapping.
EPILONT (0 – mapping)
Sindromes Epilepticos
Independent Continuant
DrugBrandName
Medication
Medication_Name
In EPSO, the AED class is present as a clinical drug component and its sub-
class drug brand name with 7-class mapping, ESSO consist of Medication_
Name class with 1-class mapping, and EPILONT consists of 0-mappping.
For Cluster III, Causes, Head Trauma is present in class mapping as follows:
Dependent Continuant
Etiology
Structural Cause
Selerosis
Buchhalter Meeting
Head Injury
EPSO ESSO
EPSO 1.000 1.000
ESSO 1.000
General Concepts
Oncology mapping
Syncope Symptom Ontology
Confusion
EPSO (0-mapping)
Occurrent
Processual Entity
Bodily Process
Physical Process
Seizure Feature
EPSO EPILONT
EPSO 1.000 0.250
EPILONT 1.000
13.5 Results
On the basis of above analysis, the computed weights are shown in Table
13.8.
Finally, the average weights of each of the 3 ontologies are computed:
It can be seen that EPSO is the suitable ontology for diagnosing epilepsy
disease.
13.7 Conclusion
This paper presents a hybrid method for ontology selection. The method
uses AHP technique to compute weights for determining the best suited
ontology for diagnosing a disease. The study demonstrates its use for diag-
nosing epilepsy. Three ontologies, EPSO, ESSO, and EPILONT are con-
sidered for the study. Focal, AED, Head Trauma, and Confused Memory
were evaluated as the important for determining the best ontology among
these three.
A Hybrid Approach to Ontology Evaluation 203
For future reference, a more improved ontology for epilepsy disease can
be developed. Ontology Evaluation and Selection can be done using other
decision making techniques.
References
1. Alobaid, A., Garijo, D., Poveda-Villalón, M., Santana-Perez, I., Fernández-
Izquierdo, A., & Corcho, O. (2019). Automating ontology engineering sup-
port activities with OnToology. Journal of Web Semantics, 57, 100472.
2. Amith, M., He, Z., Bian, J., Lossio-Ventura, J. A., & Tao, C. (2018). Assessing
the practice of biomedical ontology evaluation: Gaps and opportuni-
ties. Journal of biomedical informatics, 80, 1-13.
3. Amith, M., Manion, F., Liang, C., Harris, M., Wang, D., He, Y., & Tao, C.
(2019). Architecture and usability of OntoKeeper, an ontology evaluation
tool. BMC medical informatics and decision making, 19(4), 152.
4. Ashraf, J., Hussain, O. K., Hussain, F. K., & Chang, E. J. (2018). Representation
Phase: Ontology Usage Ontology (U Ontology). In Measuring and Analysing
the Use of Ontologies (pp. 171-203). Springer, Cham.
5. Bandeira, J., Bittencourt, I. I., Espinheira, P., & Isotani, S. (2016). FOCA: A
methodology for ontology evaluation. ArXiv preprint arXiv:1612.03353.
6. Brank, J., Grobelnik, M., & Mladenic, D. (2005, October). A survey of ontol-
ogy evaluation techniques. In Proceedings of the conference on data min-
ing and data warehouses (SiKDD 2005) (pp. 166-170). Citeseer Ljubljana,
Slovenia.
7. Fonou-Dombeu, J. V. (2019, June). A Comparative Application of Multi-
criteria Decision Making in Ontology Ranking. In International Conference
on Business Information Systems (pp. 55-69). Springer, Cham.
8. Franco, M., Vivo, J. M., Quesada-Martínez, M., Duque-Ramos, A., &
Fernández-Breis, J. T. (2019). Evaluation of ontology structural metrics
based on public repository data. Briefings in bioinformatics.
9. Haghighi, P. D., Burstein, F., Zaslavsky, A., & Arbon, P. (2013). Development and
evaluation of ontology for intelligent decision support in medical emergency
management for mass gatherings. Decision Support Systems, 54(2), 1192-1204.
10. Hooi, Y. K., Hassan, M. F., & Shariff, A. M. (2015, May). Ontology evalua-
tion—A criteria selection framework. In 2015 International Symposium on
Mathematical Sciences and Computing Research (iSMSC) (pp. 298-303).
IEEE.
11. Jarosław, W. (2018). An Attempt to Knowledge Conceptualization of Methods
and Tools Supporting Ontology Evaluation Process. Procedia computer sci-
ence, 126, 2238-2247.
12. Knoell, D., Atzmueller, M., Rieder, C., & Scherer, K. P. (2017). A Scalable
Framework for Data-Driven Ontology Evaluation. In WM (pp. 97-106).
204 Mathematics and Computer Science Volume 2
13. Lourdusamy, R., & John, A. (2018, January). A review on metrics for ontol-
ogy evaluation. In 2018 2nd International Conference on Inventive Systems
and Control (ICISC) (pp. 1415-1421). IEEE.
14. Mabkhot, M. M., Al-Samhan, A. M., & Hidri, L. (2019). An Ontology-
Enabled Case-Based Reasoning Decision Support System for Manufacturing
Process Selection. Advances in Materials Science and Engineering, 2019.
15. M’Baya, A., Laval, J., Moalla, N., Ouzrout, Y., & Bouras, A. (2016, December).
Ontology based multi criteria recomanded system to guide internship assign-
ment process. 3rd IEEE International Conference on Computational Science
and Computational Intelligence (CSCI 2016), Dec 2016, Las Vegas, United
States. ffhal-01788507f
16. Mochol, M., Oldakowski, R., & Heese, R. (2004). Ontology based recruit-
ment process. Informatik 2004, Informatik verbindet, Band 2, Beiträge der
34. Jahrestagung der Gesellschaft für Informatik eV (GI).
17. Plinere, D., & Borisov, A. (2014). Evaluation of the Ontological Knowledge
Model/Ontoloģiskā zināšanu modeļa novērtēšana/Оценка качества
онтологической модели знаний. Information Technology and Management
Science, 17(1), 81-85.
18. Saaty, T. L. (1988). What is the analytic hierarchy process? In Mathematical
models for decision support (pp. 109-121). Springer, Berlin, Heidelberg.
19. Stafstrom, C. E. (1998). The pathophysiology of epileptic seizures: a primer
for pediatricians. Pediatrics in review, 19(10), 342.
20. Sun, L., Ma, J., Zhang, Y., Dong, H., & Hussain, F. K. (2016). Cloud-FuSeR:
Fuzzy ontology and MCDM based cloud service selection. Future Generation
Computer Systems, 57, 42-55.
21. Talebpour, M., Sykora, M., & Jackson, T. (2017, November). Social and
Community Related Themes in Ontology Evaluation: Findings from an
Interview Study. In International Joint Conference on Knowledge Discovery,
Knowledge Engineering, and Knowledge Management (pp. 320-336).
Springer, Cham.
22. Verma, A. (2016, August). An abstract framework for ontology evalua-
tion. In 2016 International Conference on Data Science and Engineering
(ICDSE) (pp. 1-6). IEEE.
23. Vrandečić, D. (2009). Ontology evaluation. In Handbook on ontologies
(pp. 293-313). Springer, Berlin, Heidelberg.
24. Xu, F. X., Liu, X. H., Chen, W., Zhou, C., & Cao, B. W. (2018). An Ontology
and AHP Based Quality Evaluation Approach for Reuse Parts of End-of-Life
Construction Machinery. Mathematical Problems in Engineering, 2018.
14
Smart Health Care Waste
Segregation and Safe Disposal
R.M. Bommi1*, Sami Venkata Sai Rajeev2, Sarvepalli Navya2, Veluru Sai Teja2
and Uppala Supriya2
2
Department of CSE, Chennai Institute of Technology, Chennai, India
Abstract
Medical waste disposal has been a big issue due to an exponentially growing pop-
ulation and the COVID-19 pandemic. Increased waste generation per person
has resulted from urbanization, industrialization, and economic development.
Substandard medical waste separation at the site of origin might have a cascad-
ing effect on the environment, putting humans, wildlife, soil, and water bodies at
risk. If hazardous airborne pollutants are not effectively controlled, separated, and
burned by on-site or off-site incineration, environmental concerns linked with
inadequate clinical waste may pollute the air we breathe. This paper proposes an
IoT based smart healthcare waste segregator which segregates the waste into five
kinds. The sensors detect the type of waste and the waste gets disposed into the
smart bins accordingly. Using artificial intelligence, the status of filling of the bin is
indicated through LEDs. When the bin reaches the maximum-level, an alert mes-
sage is sent to the municipal authorities. The filled waste gets wrapped automati-
cally. The wastes which need to be incinerated is burnt in the incinerator chamber
available in the system. Therefore, this system will capably make the environment
smart, clean, and safe.
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (205–222) © 2023 Scrivener Publishing LLC
205
206 Mathematics and Computer Science Volume 2
14.1 Introduction
The Internet plays an important role in today’s world by linking com-
puters to the worldwide web (www), which permits users to access data
from anywhere in the world [1]. The Internet of Things (IoT) refers to
things that are connected to the internet and can often be managed from
there [2]. Garbage is described as solid substances generated as a result
of human activities that are removed from the system [3] because they
are no longer useful in the respective economic, biomedical, or technical
method. In a wider context, solid waste refers to all products that are
used in the home, industry, or agriculture. Municipal solid waste (MSM)
is described as waste that accrues in areas maintained by municipalities
that are responsible for its disposal and recycling. People can throw gar-
bage in waste bins, which is why they are valuable in life [3]. If this did
not happen, the future would be a mess. Because a business or household
has a garbage disposal device, it becomes a valuable piece of equipment.
The dustbin’s position as a conciliator of changing waste practices has
barely been regarded, despite its importance in our daily lives. Bins, it is
believed, are providing a telling indicator of new garbage relationships in
society as they are repurposed as environmental technologies for mod-
ern recycling schemes. Garbage and litter are all over the television these
days, with disturbing statistics of debris filling the world. Despite the
grim news, a number of people and policymakers are trying to change
the trend by creative waste management practises. The Government has
encouraged city-based schemes and public-private collaboration projects
to improve waste management systems, but these have proven to be trou-
blesome. The shortage of financial capital, sufficient expertise, and techni-
cal competencies in the public sector are the main obstacles to developing
solid waste management systems around the world. Governments have
begun to look at PPPs as a possible solution. The amount of change and
development made was minimal. Medical waste management is tough
and complicated, especially during pandemics like COVID. Due to the
apparent forte of the global outbreak, contemporary waste centers are
adapting to reveal the abnormal scientific waste and its affect on viral
propagation with particular statistics on the amount of scientific waste
generated, waste warm spots, and remedy centers. Technological know-
how [24] on inspection, segregation, transportation, storage, and reliable
waste control structures are required to optimize contemporary sources
and centers to satisfy the crisis, as healthcare waste portions are antic-
ipated to increase rapidly. For patients, healthcare workers, and waste
Smart Health Care Waste Segregation and Safe Disposal 207
The Garbage Monitoring System is a clever dustbin that makes use of sen-
sors to raise it above the trendy dustbin [13]. In the proposed system, pub-
lic trash cans would be equipped with an integrated device that allows for
real-time tracking of waste levels in the bins. The information on waste
levels would be used to create an optimal path for trash collection vans,
which will save money on fuel. The load sensors will enhance the accu-
racy of trash level information, while humidity sensors would give data on
waste segregation in a dust bin.
[14] The suggested model, which employs a convolutional neural net-
work (CNN), a prominent machine learning approach, to separate bio-
degradable and insoluble waste, is presented in this study. The plan also
includes an architectural concept for a smart garbage can that incorporates
a microprocessor and several sensors [15]. They presented an Internet of
Things-based management of solid waste in this article, which allows gar-
bage bin monitoring, dynamic scheduling, and tracking of garbage col-
lector vehicles in a city of the future. Garbage bins fitted with moderate
embedded devices are placed around the city under the proposed model.
[16] To combat COVID-19, this study suggests a leader-follower strat-
egy for hazardous infectious waste collection and government aid distribu-
tion. We suggest a smart waste material classification based on the 50-layer
residual net pre-train (ResNet-50) Convolutional Neural Network model,
which would be a machine learning tool that represents the extractor and a
Support Vector Machine (SVM) that is used to categorize waste into useful
groups/types including glass, metal, paper, and plastic among others. As
proposed by [17, 18], the “Smart Dustbin” in this article would be a cellu-
lar bin that autonomously monitors rubbish levels and transmits messages
to the appropriate city officials to update the bin’s status [19]. This article
discusses an automated system that allows depositors to dispose of their
e-waste and be compensated for it. In terms of e-waste collectors, an online
bidding session is held with the winner receiving ownership of the elec-
tronic trash that he or she won in the online bidding process. For this, an
interactive digital bin with multiple sensors and modules is required. After
user verification, the bin must be able to lock and unlock itself (Urlagunta).
The smart bin in this article is based on the Microcontroller system, which
is connected to a GSM modem and also an ultrasonic sensor. An ultra-
sonic sensor is mounted on the top of the trashcan to measure the dustbin’s
height. The minimum height is set at 10cm. The Microcontroller will be
designed such that when the trash fills, the remaining height above the
threshold height will be shown.
[20] This paper describes the design and implementation of an Internet
of Things (IoT) based Arduino microcontroller that uses ultrasonic
210 Mathematics and Computer Science Volume 2
sensors to detect the amount of trash in garbage bins, reveal the infor-
mation, update it as “empty,” “half-filled,” or “filled” on an LCD screen
at periodic intervals, and also transmit the data level information [21].
An Internet of Things (IoT) architecture for real-time garbage monitor-
ing and collection is presented in this study, with the goal of improving
and optimizing solid waste collection in cities. The Netlogo multi-agent
platform was used to mimic real-time monitoring and intelligent waste
management decisions.
[22] An automatic trash segregator is proposed in this study. When
medical waste is detected, the conveyor belt is moved by an external motor.
The trash will be sent to the sensing and classification units. The input pic-
ture is captured, pre-processed, median filtered, contrast enhanced, and
segmented in five phases [23]. The article creates the ‘Waste Hierarchy-
Technology Readiness Levels’ framework and evaluates each waste man-
agement plan against it, showing the technological maturity and the
strategy’s place in the Waste Framework directive, as well as its relative
adherence to circular economy concepts [3]. This article presents a con-
cept for a smart garbage bin surveillance system in Ethiopian cities that
uses the Global System for Mobile Communication (GSM) to organize
trash collection networks in residential and commercial sectors. The level
of waste material in the trash bin has been monitored using an ultrasonic
sensor in this suggested system and it will continually connect with the
authorized clean-up person’s cell phone via GSM modem.
This research will be crucial not only for individuals whose professions
primarily entail the disposal of garbage cans, but also for the general pub-
lic who will be following the smart city rules for a higher quality of life.
The implementation of this suggested technique would go a long way for-
ward into ensuring effective and efficient real-time garbage disposal and
resource utilization, as well as improving life for smart city residents. The
actual implementation of the functional conceptual model will serve as an
advanced as well as instructive technique of sustaining and improving the
wellbeing of smart urban residents in long run.
STEP DOWN
BRIDGE FILTER VOLTAGE
TRANSFORMER RECTIFIER CIRCUIT REGULATOR
(230 TO 12V AC) (IC 7805 & 7812)
DC MOTOR
GLASS (CONVEYOR
SENSOR BELT)
ULTRASONIC LEDS
SENSOR
MOISTURE ARDUINO
SENSOR ESP8266 MOBILE APP
PROXIMITY
SENSOR RELAY MOTOR
IoT, AI, and web/app development are the three innovations utilized in
this scheme. To control the flow of waste into the conveyor, the inlet seg-
ment has an open and close function. To detect metallic waste, an induc-
tive proximity sensor is utilized. Arduino Uno is in charge of the conveyor
belt’s pacing and rotation. An incinerator is additionally part of this con-
trivance. When a human pushes the trigger, the controls are turned off
and the waste on the conveyor belt is sent to the incinerator container.
The walls of the incinerator are composed of clay and covered with alumi-
num foil. Inside the incinerator are the ECU board electrodes and thermo-
engenderer, which are habituated to engender electricity from thermal
energy. The DC motor is driven by the engendered electricity. The sensors
are deactivated until the procedure is done. This contrivance withal has
the capability of sending an admonishment SMS if a bin is loaded. Action
is detected by an infrared sensor. Moisture sensors for dry and wet waste,
Inductive Proximity sensors for metal waste, Ultrasonic sensors for bin
filling tracking, warning messaging via Node MCU ESP8266-12E, visual
exhibit via OLED exhibit, and waste packaging via hotwire sealer are all
utilized. The Biomedical Waste Segregator is a piece of automated machin-
ery that sorts waste into four categories: metal, glass, dry, and wet. The sug-
gested framework would be capable of monitoring and managing the solid
waste amassment process as well as the total amassment process. To detect
metallic waste, an inductive proximity sensor is utilized. Dry and wet waste
is disunited utilizing a blower system. A microcontroller controls the tim-
ing and rotation of the conveyor belt. This contrivance additionally has a
feature that sends an admonition SMS if a bin is loaded.
14.3.1 Wrapping
The Ultrasonic sensor causes the door of the respective container to
close as the waste is filled in it. The container’s door is made up of two
semi-circular panels. One of the circu lar plates is connected to the other
by a thin rod, while the other is connected by an insulating rod (namely
wood). A very low volt current is passing through each rod as the doors
close. The rods (the rod with electric flow heats up) come next to each
other until the door is fully locked. The heat and compression are just
enough to melt and seal the trash container. Once the doors of the con-
tainer are closed, the current in the rod flows till the rod reaches a high
enough temperature.
Smart Health Care Waste Segregation and Safe Disposal 213
14.3.2 Incinerator
To burn the garbage, an incinerator is required. Medical waste must be
incinerated in some cases. When a human pushes the trigger, the con-
trols are turned off and the waste on the conveyor belt is sent to the
incinerator container. The walls of the incinerator are made of clay and
covered with aluminum foil. Inside the incinerator are the ECU board
electrodes and the thermos-generator, which are used to generate elec-
tricity from thermal energy. The DC motor is driven by the produced
electricity.
14.4 Methodology
The conveyor belt motor engages as the waste arrives and the conveyor
belt commences to move. Many of the motors and controls, as well as the
Smart Health Care Waste Segregation and Safe Disposal 215
Start
No Yes
Height
is full HTML page displayed
with details
No
If load is
full
Sending information to
authorized persons & server
Yes
(IoT)
Sending information to authorized
persons & server (IoT)
If the trash is not metallic, the conveyer belt sensor decides if it is a wet
or dry waste by estimating the dampness content of the waste. If there is
moisture in the waste, it is classified as wet waste and the conveyor belt is
switched off, the wet waste motor is turned on, and the garbage is depos-
ited in the wet waste container. Counter 2 is also increased. If the conveyor
belt is not carrying any wet trash, the waste is deposited into the dry waste
bin at the top of the conveyor belt. When a waste containing glass particles
is identified as glass waste, the ergo conveyor belt is turned off, the glass
waste motor is turned on, and the garbage is forced into the glass waste bin.
In consequence, Counter 3 is raised. The Figure 14.4 brings about the flow
diagram of overall process.
Conclusively, the wastes are deposited in the congruous containers,
consummating the segregation process. When the dustbin level reaches
50%, the yellow LED which is annexed to the bins gets turned on. At 75%,
the green LED gets turned on. At 90%, the red LED commences blink-
ing and at 100%, the red LED gets turned on and then the SMS alert will
be activated. When the bin gets filled, by utilizing the wrapper, the waste
gets wrapped automatically. This system has a self-cleaning system with
a conveyor belt. The conveyor chain enters the conveyor cleaner and is
Medical waste.
The sensors detect the waste and send into the particular bin.
The thermal energy which comes from the increnator will produce the electricity.
This Mobile App as seen in Figure 14.5 describes about detecting the
amount of waste in each bin. First, we need to select the state, district, and
area. Using this information, it displays the number of bins present in that
area and the amont of waste in them. As mentioned, LEDs are used for easy
identification of the amount of waste. If it is of red color, it indicates that
the bin is completely filled and needs to be cleaned. If it is yellow the bin is
partially filled and if it is green it is empty. This information is sent to the
nearest municipal corporations, so that the waste disposal becomes easy.
Declarations
Availability of Data and Materials
Information accessibility is not appropriate for this composition, as no
unused information was made or analyzed in this ponder.
Competing Interests
The Author(s) declare(s) that there is no conflict of interest.
Author’s Contribution
All the authors contributed indistinguishably to the work. All authors
examined and endorsed the ultimate composition.
References
1. B. Bharadwaj, M. Kumudha, N. Gowri Chandra, G. Chaithra, Automation of
Smart waste management using IoT to support “Swachh Bharat Abhiyan” -
A practical approach, Proc. 2017 2nd Int. Conf. Comput. Commun. Technol.
ICCCT 2017. (2017) 318–320. https://doi.org/10.1109/ICCCT2.2017.7972300.
2. A. Sharma, R. Kumar, V. Mansotra, Proposed Stemming Algorithm for
Hindi Information Retrieval, Int. J. Innov. Res. Comput. Commun. Eng.
(An ISO Certif. Organ. 3297 (2016) 11449–11455. https://doi.org/10.15680/
IJIRCCE.2016.
3. M. Karthikeyan, D. Eligo, S. Gebrehiyot, E. Bekele, D. Daniel, G. Dansa, GSM
Based Smart Waste Bin Monitoring System In Ethiopia, VI (2019) 796–804.
4. K. Pardini, J.J.P.C. Rodrigues, O. Diallo, A.K. Das, V.H.C. de Albuquerque,
S.A. Kozlov, A smart waste management solution geared towards citizens,
Sensors (Switzerland). 20 (2020) 1–15. https://doi.org/10.3390/s20082380.
5. I. Hong, S. Park, B. Lee, J. Lee, D. Jeong, S. Park, IoT-Based Smart Garbage
System for Efficient Food Waste Management, Sci. World J. 2014 (2014).
https://doi.org/10.1155/2014/646953.
220 Mathematics and Computer Science Volume 2
6. T. Anh Khoa, C.H. Phuc, P.D. Lam, L.M.B. Nhu, N.M. Trong, N.T.H. Phuong,
N. Van Dung, N. Tan-Y, H.N. Nguyen, D.N.M. Duc, Waste Management
System Using IoT-Based Machine Learning in University, Wirel. Commun.
Mob. Comput. 2020 (2020). https://doi.org/10.1155/2020/6138637.
7. K.A. Awan, I. Ud Din, A. Almogren, M. Guizani, A. Altameem, S.U. Jadoon,
RobustTrust - A Pro-Privacy Robust Distributed Trust Management
Mechanism for Internet of Things, IEEE Access. 7 (2019) 62095–62106.
https://doi.org/10.1109/ACCESS.2019.2916340.
8. V.P. Vijaynaidu, T. Dhikhi, Smart garbage management system, Int. J. Pharm.
Technol. 8 (2016) 21204–21211. https://doi.org/10.17577/ijertv4is031175.
9. C.S. Anilkumar, G. Suhas, S. Sushma, A smart dustbin using mobile appli-
cation, Int. J. Innov. Technol. Explor. Eng. 8 (2019) 3964–3967. https://doi.
org/10.35940/ijitee.K2129.0981119.
10. S. Longhi, D. Marzioni, E. Alidori, G. Di Buò, M. Prist, M. Grisostomi, M.
Pirro, Solid waste management architecture using wireless sensor network
technology, 2012 5th Int. Conf. New Technol. Mobil. Secur. - Proc. NTMS
2012 Conf. Work. (2012). https://doi.org/10.1109/NTMS.2012.6208764.
11. T.A. Khoa, M.M. Man, T.Y. Nguyen, V.D. Nguyen, N.H. Nam, Smart agricul-
ture using IoT multi-sensors: A novel watering management system, J. Sens.
Actuator Networks. 8 (2019). https://doi.org/10.3390/jsan8030045.
12. R. Asha, N. Mahajan, A. Yadav, K.S. Balamurugan, Garbage Managing Smart
System using-IOT, Int. J. Recent Technol. Eng. 8 (2019) 3646–3649. https://
doi.org/10.35940/ijrte.d7849.118419.
13. P.S.A. Mahajan, A. Kokane, A. Shewale, M. Shinde, S. Ingale, Smart Waste
Management System using IoT, Int. J. Adv. Eng. Res. Sci. 4 (2017) 93–95.
https://doi.org/10.22161/ijaers.4.4.12.
14. M.W. Rahman, R. Islam, A. Hasan, N.I. Bithi, M.M. Hasan, M.M. Rahman,
Intelligent waste management system using deep learning with IoT, J.
King Saud Univ. - Comput. Inf. Sci. (2020). https://doi.org/10.1016/j.
jksuci.2020.08.016.
15. S.S. Chaudhari, V.Y. Bhole, Solid Waste Collection as a Service using IoT-
Solution for Smart Cities, 2018 Int. Conf. Smart City Emerg. Technol.
ICSCET 2018. (2018) 1–5. https://doi.org/10.1109/ICSCET.2018.8537326.
16. J. Valizadeh, A. Hafezalkotob, S.M. Seyed Alizadeh, P. Mozafari, Hazardous
infectious waste collection and government aid distribution during COVID-
19: A robust mathematical leader-follower model approach, Sustain. Cities
Soc. 69 (2021) 102814. https://doi.org/10.1016/j.scs.2021.102814.
17. O. Adedeji, Z. Wang, Intelligent waste classification system using deep learn-
ing convolutional neural network, Procedia Manuf. 35 (2019) 607–612.
https://doi.org/10.1016/j.promfg.2019.05.086.
18. Aniqa Bano, Ikram Ud Din, Asma A. Al-Huqail, “AIoT-Based Smart Bin
for Real-Time Monitoring and Management of Solid Waste”, Scientific
Programming, vol. 2020, Article ID 6613263, 13 pages, 2020. https://doi.
org/10.1155/2020/6613263.
Smart Health Care Waste Segregation and Safe Disposal 221
19. K.B. Kumar R, S.B. Patil, S. Manjula, H.N. Kumar, An IoT Based Approach
for Efficient Collection and Disposal of E-waste, Int. J. Eng. Tech. 4 (2018)
379–383.
20. S. Abba, C. Light, IoT-Based Framework For Smart Waste Monitoring And
Control System, A Case Study of Smart Cities., (2020) 8224. https://doi.
org/10.3390/ecsa-7-08224.
21. E.D. Likotiko, D. Nyambo, J. Mwangoka, M Ulti -a Gent B Ased I Ot S Mart
W Aste M Onitoring a Nd C Ollection, International Journal of Computer
Science, Engineering and Information Technology (IJCSEIT), Vol.7, No.5,
October 2017.
22. K.G. Devi, K. Yasoda, M. Dhivya, B. Kishore, Automatic Health Care Waste
Segregation and Disposal System, J. Xidian Univ. 14 (2020) 5281–5290.
https://doi.org/10.37896/jxu14.5/573.
23. C.A. Fletcher, R. St. Clair, M. Sharmina, A framework for assessing the cir-
cularity and technological maturity of plastic waste management strategies
in hospitals, J. Clean. Prod. 306 (2021) 127169. https://doi.org/10.1016/j.
jclepro.2021.127169.
24. Sarkodie, S.A., Owusu, P.A. Global assessment of environment, health
and economic impact of the novel coronavirus (COVID-19). Environ Dev
Sustain 23, 5005–5015 (2021). https://doi.org/10.1007/s10668-020-00801-2.
15
Investigation of Viscoelastic
Magnetohydrodynamics (MHD)
Flow Over an Expanded Lamina
Surrounded in a Permeable Media
Hiranmoy Mondal1*, Arindam Sarkar2 and Raj Nandkeolyar2
1
Department of Applied Science, Maulana Abul Kalam Azad University of
Technology, West Bengal, Kolkata, India
2
Department of Mathematics, National Institute of Technology Jamshedpur,
Jamshedpur, India
Abstract
The results of the boundary value problem of MHD nanofluid micropolar fluid
flow have been carried out throughout this study. Using the spectral quasi-
linearization method (SQLM), the mass as well as heat transfer effects of fluid have
been investigated. A stream consists of viscous dissipation, bouncy force, thermal
radiation, and Joule heating. It has been scrutinized that the existence of magnetic
parameter Mn enhances concentration and temperature gradient, whereas the
appearance of permeable media turns down the velocity profile.
15.1 Introduction
15.1.1 Literature Review
Viscoelastic Magneto-hydrodynamic nanofluids through a stretching
sheet have a broad range of practical applications in industries such as
paper production, metal spinning, relegation of plastic lamina, continuous
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (223–238) © 2023 Scrivener Publishing LLC
223
224 Mathematics and Computer Science Volume 2
15.1.2 Nomenclature
B0 Strength of uniform magnetic field
Cp Specific heat of the fluid
f Stream function
K Thermal conductivity of the fluid
Mn Magnetic parameter
R Radiation parameter
T Non-dimensional temperature
T∞ Ambient temperature/temperature far from sheet
ρ Density of the fluid
Kp Porosity parameter
k0 Dimensionless elastic parameter
K ′p Porosity parameter
s Plate concentration parameter
r Plate temperature parameter
D Molecular diffusivity
DB Brownian motion coefficient
DT Thermophoretic diffusion coefficient
µ Dynamic viscosity
ν Kinematic coefficient of viscosity
σ Electrical conductivity
Rc Elastic parameter
Pr Prandtl number
Viscoelastic MHD Flow Over an Expanded Lamina 225
Nt Thermophoresis parameter
Nb Brownian motion parameter
Le Lewis number
T∞ C∞
Thermal Boundary Layer
y
Porous Medium
Tw Cw
∂T ∂T ∂ 2T ∂T ∂C DT ∂T 2
ρC p u +v = k 2 + τ DB + (15.3)
∂x ∂y ∂y ∂y ∂y T∞ ∂y
∂C ∂C ∂ 2C D ∂ 2T (15.4)
u +v = D 2 + T 2 − R(C − C∞ )
∂x ∂y ∂y T∞ ∂y
∂Ψ ∂Ψ
u= ,v = − . (15.6)
∂x ∂y
λ
=η y= , Ψ (x , y ) x vλ f (η )
ν
Substituting all the above values, Equations (15.2), (15.3), and (15.4) are
transformed as follows:
1
f ′′′ + ff ′′ − f ′2 − Rc{2 f ′ f ′′′ − ff ′′2 − ff iv } − Mn + f ′ = 0 (15.7)
KP
1 1 Nt (15.9)
sf ′∅ − f ∅′ = ∅′′ + θ ′′ − R1∅
Le Le Lb
θ(0) = 1, θ(∞) = 0
1
F ≡ f ′′′ + ff ′′ − f ′2 − Rc{2 f ′f ′′′ − f ′′2 − ff iv } − Mn + f′=0
KP
∂F
a0, r = iv = Rcf r
∂f
∂F
a1, r = = 1 − 2Rcf r′
∂f ′′′
∂F
a2, r = = f r + 2Rcf r′′
∂f ′′
∂F 1
a3, r = = −2 f r′ − 2Rcf r′′′− Mn +
∂f ′ K P
∂F
a4 , r = = f r′′+ Rcf riv
∂f
b0, rθr′′+1 + b1, rθr′+1 + b2, rθr +1 + b3, r f r′+1 + b 4, r f r +1 + b5, r ∅′r +1 = Rθ
Nt
∅ ≡ ∅′′ + θ ′′ − LeR1∅ − Le(sf ′∅ − f ∅′)
Nb
∂∅
c0, r = =1
∂∅′′
∂∅
c1, r = = Lef r
∂∅′
∂∅
c2, r = = −LeR1 − sLe f r′
∂∅
∂∅
c 3, r = = −sLeθr
∂f ′
∂∅
c 4, r = = Le∅′r
∂f
∂∅ N b
c 5, r = =
∂θ ′′ N t
Table 15.1 Values (Cf), (Nu), and (Sh) for different parameters.
Mn Sr R1 Nb Nt −f″(0) −θ′(0) −Ø′(0)
0.0 0.7 0.2 0.5 0.2 6.47747836 -1.24792674 -0.16130247
0.2 6.68965205 -1.23881769 -0.15872752
0.6 7.09600494 -1.22067600 -0.15482334
0.2 0.7 0.2 0.5 0.2 6.68965205 -1.23881769 -0.15872752
1.6 6.68959350 -2.06833985 0.15987681
5.0 6.69009405 -4.33945643 1.04495979
0.2 0.7 0.0 0.5 0.2 6.68959327 -1.57062405 0.21775166
0.4 6.69037312 -1.09209497 -0.36582335
0.6 6.69002619 -1.00077201 -0.51685468
0.2 0.7 0.2 0.2 0.2 6.68988106 -1.67046286 0.80664545
0.6 6.68993581 -1.12522670 -0.25586930
0.8 6.69012545 -0.93800745 -0.36733042
0.2 0.7 0.2 0.5 0.1 6.68959286 -1.32703668 -0.32973517
0.4 6.68990575 -1.09355804 0.11126215
0.7 6.69039059 -0.93732032 0.40484395
0.8
Rc = 0.1; Kp = 0.5; Sr = 0.2; S = 0.5;
Pr = 8; Nb = 0.5; Nt = 0.2;
Le = 0.6; R1 = 0.5;
0.6
Mn = 0.1
f’(η)
Mn = 2.5
Mn = 3.5
0.4
Mn = 4.7
0.2
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
η
0.6
0.5
0.4
0.3
f(η)
-0.1
0 1 2 3 4 5 6 7 8 9 10
η
0.6 Mn = 0.1
Mn = 2.5
Φ(η)
0.4 Mn = 3.5
Mn = 4.7
0.2
-0.2
0 1 2 3 4 5 6 7 8 9 10
η
0.8
Rc = 0.1; Kp = 0.5; Sr = 0.2; S = 0.5;
Pr = 8; Nb = 0.5; Nt = 0.2; Le = 0.6; Che = 0.5
0.6 Mn = 0.1
θ(η)
Mn = 2.5
Mn = 3.5
0.4 Mn = 4.7
0.2
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
η
0.8
Mn = 2; Kp = 0.5; Sr = 0.2; S = 0.5
Rc = 0.1; Nb = 0.5; Nt = 0.2; Le = 0.6; Che = 0.5
0.6
θ(η)
Pr = 6.8
Pr = 8.9
0.4 Pr = 10
Pr = 12
0.2
0
0 1 2 3 4 5 6 7 8 9 10
η
0.5
0.4
Mn = 2; Kp = 0.5; Sr = 0.2; S = 0.5;
0.3 Pr = 8; Nb = 0.5; Nt = 0.2; Le = 0.6; Che = 0.5;
Rc = 0
f(η)
0.2 Rc = 0.01
Rc = 0.1
0.1 Rc = 0.2
-0.1
0 1 2 3 4 5 6 7 8 9 10
η
0.8
Mn = 2; Kp = 0.5; Sr = 0.2; S = 0.5;
Pr = 8; Nb = 0.5; Nt = 0.2; Le = 0.6; Rc = 0.1
0.6
Φ(η)
Che = 0.5
Che = 2
0.4 Che = 5
Che = 8
0.2
0
0 1 2 3 4 5 6 7 8 9 10
η
0.8
Mn = 2; Kp = 0.5; Sr = 0.2
0.6 Pr = 8; Nb = 0.5; Nt = 0.2;
Le = 0.6; Rc = 0.1; Che 0.5
S = 0.5
Φ(η)
0.4 S=2
S=8
0.2 S = 10
-0.2
0 1 2 3 4 5 6 7 8 9 10
η
Figure 15.9 Results of plate concentration on Ø(η) for Mn = 2, Rc = 0.1, Che = 0.5.
Viscoelastic MHD Flow Over an Expanded Lamina 235
0.8
Mn = 2; Kp = 0.5; Sr = 0.2;
Pr = 8; Nb = 0.5; Nt = 0.2;
0.6 Le = 0.6; Rc = 0.1; Che = 0.5, S = 0.5
Le = 0.6
Φ(η)
0.4 Le = 1
Le = 2
Le = 5
0.2
-0.2
0 1 2 3 4 5 6 7 8 9 10
η
Figure 15.10 Results of Lewis parameter Le on Ø(η) for Mn = 2, Rc = 0.1, Che = 0.5.
0.6
0.5
0.4
0.3 Mn = 2; Sr = 0.2; Pr = 8; Nb = 0.5; Nt = 0.2;
f(η)
0.2
Kp = 0.5
Kp = 1
0.1
Kp = 1.6
Kp = 2.5
0
-0.1
0 1 2 3 4 5 6 7 8 9 10
η
at each point and after a certain time, transverse velocity becomes parallel
to the axis of η.
15.4 Conclusion
The solutions for a steady boundary layer stream and heat distribution for
a nanofluid throughout an exponentially expanded lamina in the presence
of a chemical reaction is analysed. Results of the magnetic parameter and
the viscoelastic parameter on various velocities are discussed. The impres-
sion of the magnetic parameter, Prandtl number, and the heat sink/source
parameter on the temperature profiles are presented. The main findings of
this research may be shortened as follows:
236 Mathematics and Computer Science Volume 2
References
1. D. Pal and H. Mondal, Influence of thermophoresis and Soret-Dufour on
magnetohydrodynamic heat and mass transfer over a non-isothermal wedge
with thermal radiation and Ohmic dissipation, J. Magn. Magn. Mater., 331,
250–255, 2013.
2. D. Pal and H. Mondal, Influence of Soret and Dufour on MHD buoyancy-
driven heat and mass transfer over a stretching sheet in porous media with
temperature-dependent viscosity, Nucl. Eng. Des., 256, 350–357, 1970.
3. S. Baag, S. R. Mishra, G. C. Dash, and M. R. Acharya, Entropy generation
analysis for viscoelastic MHD flow over a stretching sheet embedded in a
porous medium, Ain Shams Eng. J., 8, no. 4, 623–632, 2017.
4. D. Pal and H. Mondal, Influence of temperature-dependent viscosity and
thermal radiation on MHD forced convection over a non-isothermal wedge,
Appl. Math. Comput., 212, no. 1, 194–208, 2009.
5. Crane LJ. Flow past a stretching plane. Jape Math Phys (ZAMP), 21, 645–7,
1970.
6. Ishak A, Nazar R, Pop I. Heat transfer over a stretching surface with variable
heat flux in micropolar fluids. Phys Lett A, 372, 559–61, 2008.
7. Ishak A, Nazar R, Arifin NM, Pop I. Mixed convection of the stagnation
point flow towards a stretching vertical permeable sheet. Malaysian J Math
Sci 2, 217–226, 2007.
8. Ishak A, Nazar R, Pop I. Mixed convection boundary layers in the stagna-
tion point flow towards a stretching vertical permeable sheet. Mechanica 41,
509–518, 2006.
9. Yao S, Fang T, Zhong Y. Heat transfer of a generalized stretching/shrinking
wall problem with convection boundary conditions. Commun Nonlinear Sci
Numer Simul 16, 752–760, 2011.
10. Quasim M, Hayat T, Obaidat S. Radiation effect on the mixed convection
flow of a viscoelastic fluid along an inclined stretching sheet. Z Naturforsch
67, 195–202, 2012.
11. Hayat T, Quasim M, Mesloub S. MHD flow and heat transfer over a stretch-
ing sheet with Newtonian heating with slip condition. Int J Numer Methods
Fluid 66, 963–975, 2011.
Viscoelastic MHD Flow Over an Expanded Lamina 237
12. Hayat, T., Qayyum, S., Alsaedi, A. and Ahmad, B., Magnetohydrodynamic
(MHD) nonlinear convective flow of Walters-B nanofluid over a nonlinear
stretching sheet with variable thickness, International Journal of Heat and
Mass Transfer, 110, 506-514, 2017.
13. Daniel, Y.S., Aziz, Z.A., Ismail, Z. and Salah, F., Double stratification effects
on unsteady electrical MHD mixed convection flow of nanofluid with viscous
dissipation and Joule heating, Journal of Applied Research and Technology,
15(5), 464-476, 2017.
14. Hayat, T., Qayyum, S., Khan, M.I. and Alsaedi, A., Entropy generation in
magnetohydrodynamic radiative flow due to rotating disk in presence of
viscous dissipation and Joule heating, Physics of Fluids, 30(1), 017101,
2018.
15. Hayat, T., Qayyum, S., Alsaedi, A. and Shafiq, A., Inclined magnetic field and
heat source/sink aspects in flow of nanofluid with nonlinear thermal radia-
tion, International Journal of Heat and Mass Transfer, 103, 99-107, 2016.
16. Kishan, N. and Maripala, S., Thermophoresis and viscous dissipation effects
on Darcy-Forchheimer MHD mixed convection in a fluid saturated porous
media, Advances in Applied Science Research, 3(1), 60-74, 2012.
17. Ibrahim, W., Shankar, B. and Nandeppanavar, M.M., MHD stagnation point
flow and heat transfer due to nanofluid towards a stretching sheet, Int. J. Heat
Mass Transfer., 56, 1-9, 2013.
18. Turkyilmazoglu, M., Exact analytical solutions for heat and mass transfer of
MHD slip flow in nanofluids, Chem. Eng. Sci., 84, 182-187, 2012.
19. Farooq, M., Ijaz Khan, M., Waqas, M., Hayat, T., Alsaedi, A. and Imran Khan,
M., MHD stagnation point flow of viscoelastic nanofluid with non-linear
radiation effects, Journal of Molecular Liquids., 221, 1097-1103, 2016.
20. Sheikholeslami, M., Hari Kataria, R. and Akhil Mittal, S., Effect of thermal
diffusion and heat-generation on MHD nanofluid flow past an oscillating
vertical plate through porous medium, Journal of Molecular Liquids, 257,
12-25, 2018.
21. Baag, S., Mishra, S. R., Dash, G. C. and Acharya, M. R., Entropy generation
analysis for viscoelastic MHD flow over a stretching sheet embedded in a
porous medium, Ain Shams Engineering Journal, 8(4), 623-632, 2017.
22. Sheikholeslami, M. and Rokni, H. B., Influence of melting surface on MHD
nanofluid flow by means of two-phase model, Chinese J. Physics, 55(4),
1352-1360, 2017.
23. Abdul Hakeem, A.K., Govindaraju, M. and Ganga, B., Influence of inclined
Lorentz forces on entropy generation analysis for viscoelastic fluid over a
stretching sheet with nonlinear thermal radiation and heat source/sink,
J. Heat Mass Transfer Research., 6(1), 1-10, 2019.
24. Emad M. Abo-Eladahab, Mohamed A. El Aziz, Blowing/suction effect on
hydromagnetic heat transfer by mixed convection from an inclined contin-
uously stretching surface with internal heat generation/absorption, Int. J.
Therm. Sci., 43, 709-719, 2004.
238 Mathematics and Computer Science Volume 2
25. Subhas Abel, M. and Mahesha, N., Heat transfer in MHD viscoelastic fluid
flow over a stretching sheet with variable thermal conductivity, non-uniform
heat source and radiation, Appl. Math. Model., 32, 1965-1983, 2008.
26. Ganga, B., Govindaraju, M. and Abdul Hakeem, A.K., Effects of inclined
magnetic field on entropy generation in nanofluid over a stretching sheet
with partial slip and nonlinear thermal radiation, Iranian Journal of Science
and Technology: Transactions of Mechanical Engineering, 1139, 2018.
27. Zeeshan A, Majeed A. Effect of magnetic dipole on radiative non-Darcian
mixed convective flow over a stretching sheet in porous medium. J Nanofluids
4, 617–626, 2016.
28. H Mondal, S Ghosh, PK Roy, S Chatterjee, Effects of Ion-Slip and Hall
Currents on Magnetohydrodynamic Nanofluid Flow with Thermal Diffusion
Using Spectral Quasi-Linearization Method, Journal of Nanofluids 10 (4),
608-615.
29. S. S. Motsa. A new spectral local linearization method for nonlinear bound-
ary layer flow problems. Journal of Applied Mathematics, 2013.
16
Quickest Multi-Commodity Contraflow
with Non-Symmetric Traversal Times
Shiva Prakash Gupta1*, Urmila Pyakurel2 and Tanka Nath Dhamala2
2
Central Department of Mathematics, Tribhuvan University, Kathmandu, Nepal
Abstract
Routing multiple distinct commodities from particular supply points to appro-
priate destination points via the arcs of a network topology while preserv-
ing capacity limits is the most challenging issue in operations research. The
quickest multi-commodity flow issue reduces the delivery time to complete
the process. Computationally, this problem is NP-hard. We consider asymmet-
ric transit times on anti-parallel lanes due to uneven road conditions and flow
dependency. Outbound lane capacity can be boosted by reorienting lanes towards
demand locations. We incorporate the contraflow strategy to the quickest multi-
commodity flow problem and provide an approximation scheme utilizing length-
bounded flow with non-symmetric traversal times on oppositely oriented lanes.
16.1 Introduction
The topology of a transport system is described as a network that cor-
relates to the transshipment of diverse commodities, with distribution
centers, demand centers, and intersections of road segments serving as
nodes and arcs serving as linkages between nodes. The starting and end-
ing destinations of commodities are known as supply points and demand
points, respectively. Flow refers to the collection of goods that have been
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (239–250) © 2023 Scrivener Publishing LLC
239
240 Mathematics and Computer Science Volume 2
Lozovanu et al. [14] and Kappmeir [15] used a time expansion approach
for solving the maximum dynamic multi-commodity flow problem.
This approach is extended to multiple sources and a single sink multi-
commodity earliest arrival transshipment in [15].
Contraflow refers to the switching arc orientations to boost traffic flow
value and minimize traversal time by increasing the capacity of lanes in
a network. For a single-source and a single-sink, the authors in [16] pro-
posed models and strongly polynomial-time methods to solve maximum
and quickest flow problems. At time zero, these contraflows are made and
then rectified. Nath et al. [17] investigated the problem with asymmetric
traversal times on oppositely directed lanes and modified the algorithm of
[16] so that traversal time of a non-reversed arc is taken after lane reversals
and solved the problem in polynomial-time for a single-commodities. In
[18], the authors reported a polynomial-time solution to the earliest arrival
transshipment problem and the maximum flow problem with priority
on terminals. Gupta et al. [19] investigated the non-polynomial solution
of the generalized maximum flow issue on a lossy network with non-
symmetric traversal times on anti-parallel lanes. Recently, the authors in
[20] extended this to the case of multi-commodity with symmetric tran-
sit times and the same authors in [21] provided FPTAS to a maximum
dynamic multi-commodity contraflow problem with non-symmetric tra-
versal times on oppositely oriented lanes.
The fundamental objective of partial contraflow is to use the capacity
of underutilized lanes and was introduced by Pyakurel et al. [22]. In times
of emergency, the stored capacity of unoccupied arcs might be utilized for
logistical assistance and facility placement.
We present the quickest multi-commodity contraflow (QMCCF) prob-
lem with non-symmetric traversal times on anti-parallel lanes in this paper.
Fleischer and Skutella [11, 13] developed an approximation algorithm to
this problem without contraflow using a T-length bounded function.
The authors in [23, 24] introduced a partial lane reversal approach
with symmetric traversal times on oppositely directed lanes. We provide
the polynomial-time approximate solution of the problem by utilizing a
T-length bound approximation. The travel time for transshipping com-
modities from origin nodes to destination nodes is reduced by using the
lane reversal strategy in the routing problem. The study’s main relevance is
the decrease in delivery time.
The rest of the article is arranged as follows. Section 16.2 describes fun-
damental concepts with flow models. The QMCCF problem with non-
symmetric traversal times on oppositely oriented lanes is introduced in
242 Mathematics and Computer Science Volume 2
the backward arc of e. The capacity is re-defined as the total of the capaci-
ties of the anti-parallel lanes, in Na with ue = 0, if e ∉A. The traversal time
τa along with arc a in Na is the same as the traversal time of non-reversed
arcs. For a single lane, we assume τ a = τ e = τ er .. Other network parameters
are unaltered.
A static network is one that lacks a time dimension and is defined by
N = (V, A, K, u, di, S+, S−).
The static flow function is defined by x : A → R+. To compute the solu-
tion of real world dynamic flow problems, many useful properties of static
network topology are essential tools. A multi-commodity flow over time
Φ in a network N with contraflow is a collection of commodities described
by Φi : Aa × T → R+ meeting the requirements (16.2−16.4).
min T (16.1)
Multi-commodity Contraflow with Non-symmetric Traversal Times 243
subject to,
T T di if v = si
∑∑ +
θ =0 e∈δ (v )
i
Φ (θ ) −
e ∑∑
−
θ =τ e e∈δ (v )
i
Φ (θ − τ e ) = − di if=
ev = ti ∀i ∈ K
0 otherwise
(16.2)
σ σ
∑∑
θ =0 e∈δ + (v )
Φie (θ ) − ∑ ∑ Φ (θ −τ ) ≤ 0, ∀v ∉ {s ,t },i ∈ K ,σ ∈T
θ =τ e e∈δ − (v )
i
e e i i
(16.3)
∑
0 ≤ Φ e (σ ) =Φie (σ ) ≤ ue + uer , ∀e ∈ Aa , i ∈ K ,σ ∈ T (16.4)
e∈δ + (v )
In this case, the third equation in constraint (16.2) denotes flow con-
servation restrictions at time T, but the constraints in (16.3) denote non-
conservation needs at intermediate time points T. Likewise, capacity with
lane reversals limits the bundle restrictions in (16.4). The purpose is to
deliver a predetermined quantity of flow to fulfill the demand for each
commodity from supply points to destination locations in the least amount
of time, as specified by the first two conditions of (16.2).
The cost is defined by
T
c(Φ) = ∑ ∑ ∑ Φ (θ )
e∈A i∈K
cei
θ=
0
i
e (16.5)
∑ ∑ Φ (θ ) ≤ C
e∈A
cei
θ=
0
i
e i (16.6)
s1 3,1 v t2 s1 3,1 v t2
s2 6,1 w t1 s2 6,1 w t1
(a) (b)
Figure 16.1 (a) Network with multi-commodity having capacity and traversal time on
arcs. (b) Arc reversals with symmetric and non-symmetric traversal times.
are used to transport the first and second commodities, respectively. A contra-
flow network is depicted in Figure 16.1(b).
Authors in [25] modeled the quickest contraflow problem for a single
commodity as an integer programming problem. Furthermore, they pro-
posed greedy and bottle-neck relaxation heuristics to provide a numeri-
cal solution. Rebennack et al. [16] presented its polynomial-time solution.
However, the problem with numerous sources and/or sinks is NP-complete
via a reduction from 3-SAT [16]. In [26, 27], the continuous-time variant
of a single-source single-sink maximum and the quickest contraflow prob-
lems are polynomially solvable. The authors in [22] introduced and solved
the problem with partial lane reversals.
∑
P ∈Pi
with x Pi > 0 and the length τ P = τ e of any path P ∈ Pi is at most T.
e ∈P
A T-length, restricted static flow x that matches multi-commodity con-
straints is NP-hard to discover [13]. This issue, on the other hand, is
polynomial-time approximated.
We compute an approximate solution to the QMCCF problem with
non-symmetric traversal times on anti-parallel lanes by applying length-
bounded flows as described [23, 24] and presented in Algorithm 1. In Step
2 of our approach, we first review the solution technique of [11, 13]. All
246 Mathematics and Computer Science Volume 2
other steps of the algorithm are feasible according to [23]. The removal of
the cycle flow in Step 3 suggests that flow only travels in one direction, not
both. In Step 5, the capacity of the unutilized arcs is preserved. We argue
that a possible approximate solution to the QMCCF with non-symmetric
traversal time on oppositely oriented lanes on network N would likewise
be a feasible approximate solution to the quickest multi-commodity flow
issue on modified network Na.
The authors in [17] have shown that the complexity of the lane rever-
sal problems having non-symmetric travel times on oppositely oriented
lanes is the same as symmetric travel time on oppositely oriented lanes.
As a result, an approximate QMCCF with non-symmetric traversal time
on oppositely oriented lanes may be optimally calculated on network N.
Hence, we have Theorem 2.
u=
a ue + uer ,,
τ e if arc e re is
if arc r reverted towards e
is reverted towards e
τ a : = τa: =
τ er if arc e eisreverted
if arc reverteetowards er er .
d towards
16.4 Conclusions
One of the primary issues in operations research is routing many com-
modities from their delivery points to target points over a shared network.
A crucial concern is the reduction of time (cost). A well-known quickest
flow problem was addressed to deliver the requisite demand in the shortest
feasible period. For a single-commodity, this issue was solved in strongly
polynomial-time, but for several commodities it is NP-hard. However, a
polynomial-time approximation technique based on a length-bounded
function was found. The lane reversals technique is an essential tool for
improving the quickest time in the two-way network. We incorporated this
approach in length-bounded approximation.
In this work, we investigated the QMCCF problem with non-symmetric
traversal times on oppositely oriented lanes. We presented its mathemati-
cal model and provided a polynomial-time approximation algorithm. This
study minimizes the routing time significantly. By analyzing the results
of the previously studied problem with constant transit time, it would be
interesting to extend these strategies to flow-dependent transit times. The
findings in this research have both theoretical and practical implications.
Acknowledgments
The first author wishes to thank the University Grant Commission of
Nepal for the Ph.D. research grant and the second author wishes to thank
the Alexander Von Humboldt Foundation for the Digital Cooperation
Fellowship (August 1, 2021 to January 31, 2022).
References
1. R. K. Ahuja, T. L. Mangati, and J. B. Orlin, Network flows: theory, algorithm,
and applications, Prentice Hall, Englewood Cliffs, 1993.
Multi-commodity Contraflow with Non-symmetric Traversal Times 249
Abstract
A production inventory model for a perishable good decaying at a constant rate
has been devised in this research article. A ramp-type demand function has been
considered, that is, the demand follows an accelerated growth initially for some
time and finally becomes constant. The model allows shortage and is completely
backlogged. The cost of holding the inventory is considered to vary linearly with
time. Four cases have been developed depending on the position of time point
at which the demand becomes constant in a trapezoidal-type demand. The pro-
duction cycle restarts after a certain time. Optimal production stopping time and
resuming time are calculated to optimize the expense of the production set-up per
unit. The model is illustrated with a numerical.
17.1 Introduction
Demand is always fluctuating according to the requirements and prefer-
ences of customers and is quite unpredictable. In case of new launches
of items like cosmetics, electronic gadgets with new technologies, trendy
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (251–264) © 2023 Scrivener Publishing LLC
251
252 Mathematics and Computer Science Volume 2
garments, etc. in the market, the items are in high demand initially. But,
with the passage of time as more improvised items come into the market,
the demand for the old ones become constant and stabilized. In order to
deal with this kind of demand trend, trapezoidal demand has been dealt
with in various studies since Ritchie (1980) [1]. Shaikh et al. (2020) [2]
devised a stock optimization model for perishing goods with a preser-
vation facility and a trapezoidal demand. Yang (2019) [3] developed an
optimization model for a trapezoidal order trend having bi-level business
credit type financing while placing orders. Chauhan et al. (2021) [4] devel-
oped an EOQ model for a trapezoidal type ordering system. During the
stock out or shortage duration, backlogging is considered, also allowing
certain discount strategies. Sana (2010) [5] considered deterioration and
partial backlogging. Kawakatsu (2010) [6] devised an optimal refilling
strategy for a retailer. A seasonal good with a trapezoidal order trend is
considered with focus on Special Display Goods. Avinadav et al. (2013) [7]
have developed a model where the demand varies with price of the item.
Wang and Huang (2014) [8] devised a model for periodic items having
price and trapezoidal-type order trends and deterioration. Ahmed et al.
(2013) [9] proposed a fresh method for investigating the EOQ (Economic
Order Quantity) assuming a trapezoidal demand, partial backlogging, and
a general perish rate.
Perishing of items is an unavoidable phenomenon that should not be
neglected while formulating an inventory system. But, it definitely depends
on the kind of item taken into consideration. For example, for highly
deteriorating items like fruits, vegetables, etc., deteriorating items should
be taken as a variable function of time, quantity, etc. But, items such as
metals (steel utensils, iron rods, etc.), medicines (expiry), etc. deteriorate
extremely slow and hence the rate of deterioration might be considered
constant. In this study, the items being studied are of the non-depreciat-
ing type like electronic goods, medicines, COVID protective gears, etc.
In recent time, Halima et al. (2021) [10] devised an overtime production
model for goods with a constant rate of perish and the demand depend-
ing on non-linear price and stock dependent order trend. Mandal and Pal
(1998) [11] have developed inventory models for perishing goods that fol-
lows a trapezoidal demand.
Shortages of items are very common in the market system. When short-
ages occur, it is completely a customer’s will whether to wait for the back-
order or to go to some other seller for instant availability. Backlogging
rate varies with the holdback time for the next refilling. Goyal (1988) [12]
established a model for replenishment of trendy/fashionable inventories
including shortages. Giri et al. (2000) [13] devised a model on lot sizing
Mathematical Representation for Deteriorating Goods 253
study for perishing goods with stockout situations and demand depending
on time. Wang (2002) [14] devised a replenishment strategy for perishing
items with inventory running with shortages and thus, partial backlog-
ging. Ghosh and Chaudhuri (2005) [15] developed an economic order
quantity model for a perishable good with trended demand and stockout
situations in each cycle. Pentico and Drake (2011) [16] did a study on
deterministic models for the economic order quantity models and eco-
nomic production quantity models considering backordering. Sharma
and Singh (2013) [17] devised a replica for perishable goods, accounting
for consumption rate and backlogging. Roy Chowdhury et al. (2014) [18]
designed an order-level inventory model for a decaying good having a
quadratic type ordering pattern and backlogging during a stock out situ-
ation in every cycle. Tripathi (2015) [19] developed a model for perishing
goods with an escalating order trend and stock out situation under infla-
tionary spiral.
In the present work, a production-inventory model for a constantly per-
ishable good with a trapezoidal-type order trend has been considered. The
production starts at an initial time and is proportional to demand. The cost
of holding inventory varies linearly with time. Shortage is allowed and is
completely backlogged. Three cases have been dealt with depending on the
position of the time at which the trapezoidal-type demand gains stability.
We have illustrated the model numerically.
17.2.2 Notation
A: The unit ordering cost per cycle ($/unit)
P (t): Production rate varies directly with demand D(t), that is, P (t) =
βD(t); this means that the higher the demand, the greater the production;
this is required for ramp-type demand as initially demand is more, hence
production should also be more to satisfy the customer demand, but the
demand stabilizes after sometime and the production has to be reduced
accordingly in order to avoid wastage of items and loss
ν: The parameter of the trapezoidal demand function (time point) where
demand becomes constant
D(t): In case of certain items like COVID gear, with the outbreak of
COVID-19 in India in the year 2020, the demand of masks and PPEs saw
an increasing demand in the beginning but after some time the demand
became stabilized. This kind of demand is a ramp type demand which can
be mathematically expressed as: Demand at the time t ≥ 0
H (t − v ) = 1 (t ≥ ν )
(17.2)
= 0, (t < ν )
dI1A (t )
= (β − 1)(a + bt ) − θ I1A (t ), (0 ≤ t ≤ ν )
dt
dI1B (t )
= (β − 1)(a + bν ) − θ I1B (t ), (ν ≤ t ≤ t a )
dt
dI1C (t )
= −(a + bν ) − θ I1C (t ), (t a ≤ t ≤ tb )
dt
dI1D (tt )
= −(a + bν ), (tb ≤ t ≤ t c )
dt
dI1E (t )
= (β − 1)(a + bν ), (t c ≤ t ≤ T ) (17.3)
dt
CH= (h + st )Z0v I1A (t )dt + (h + st )Zvta I1B (t )dt + (h + st )Zttab I1C (t )dt
(17.5)
Mathematical Representation for Deteriorating Goods 257
3. Shortage Cost
C=
S {
cs −Zttbc I1D (t )dt − ZtTc I1E (t )dt } (17.6)
4. Production Cost
C=
P c p {−Z0v β (a + bt )dt + Zvta β (a + bν )dt + ZtTc β (a + bν )dt} (17.7)
dI 2 A (t )
= (β − 1)(a + bt ) − θ I 2 A (t ), (0 ≤ t ≤ t a )
dt
dI 2 B (t )
= −(a + bt ) − θ I 2 B (t ), (t a ≤ t ≤ ν )
dt
dI 2C (t )
= −(a + bν ) − θ I 2C (t ), (ν ≤ t ≤ tb )
dt
dI 2 D (t )
= −(a + bν ),
= (tb ≤ t ≤ t c )
dt
dI 2 E (t )
= (β − 1)(a + bν ), (t c ≤ t ≤ T ) (17.9)
dt
(β − 1)(aθ + b(−1 + tθ ))
I 2 A (t ) =
θ2 (0 ≤ t ≤ t a )
(β − 1)(aθ − b)
−e −tθ ,
θ2
(aθ + b(−1 + tθ ))
I 2 B (t ) = −
θ2
{
+ e −tθ
(aθ + b(−1 + t aθ )) taθ (β − 1)(aθ − b)
θ2
βe −
θ2
, } (t a ≤ t ≤ ν )
(a + bν ) (tb−t )θ
I 2C (t ) = (e − 1), (ν ≤ t ≤ tb )
θ
I 2 D (t ) = (a + bν )(tb − t ), (tb ≤ t ≤ t c )
I 2 E (t=) (a + bβ )(β − 1)(t − T ), (t c ≤ t ≤ T )
(17.10)
3. Shortage Cost
C=
S {
cs −Zttbc I 2 D (t )dt − ZtTc I 2 E (t )dt } (17.12)
4. Production Cost
C=
P c p {−Z0td β (a + bt )dt + ZtcT β (a + bν )dt} (17.13)
dI 3 A (t )
= (β − 1)(a + bt ) − θ I 3 A (t ), (0 ≤ t ≤ t a )
dt
dI 3 B (t )
= −(a + bt ) − θ I 3 B (t ), (t a ≤ t ≤ tb )
dt
dI 3C (t )
= −(a + bt ), (tb ≤ t ≤ ν )
dt
dI 3D (t )
= −(a + bν ), (ν ≤ t ≤ t c )
dt
dI 3 E (t )
= (β − 1)(a + bν ), (t c ≤ t ≤ T ) (17.15)
dt
3. Shortage Cost
∫ {a(t − t ) + 2 (t − t )}dt
v
b 2 2
C=
s cs {− b b
tb
− {−(a + bv )t + at + (t + v )}dt
tc
b
∫ v 2
b b
2 2
(17.18)
T
∫
− {(a + bv )(β − 1)(t − T )}dt }
tc
4. Production Cost
{
CP = c p − J t0a β (a + bt )dt + J T β (a + bν )dt
tc } (17.19)
{
CD = cd ( β − 1) J t0a (a + bt )dt − J tb (a + bt )dt
td } (17.20)
The average total system expenditure for the above three instances is
given by:
TC = T1 [CO + CH + CS + CD + CP ]
The total system cost TC depends on three variables ta, tb, and tc. We
need to find the optimized values of ta, tb, and tc that is, tm, tm, and tm respec-
tively, that minimize the average total system cost.
∂TC ∂TC ∂TC
tm, tm, and tm are the solution of = 0, = 0, = 0, respectively,
∂t1 ∂t 2 ∂t 3
which are the necessary conditions for the optima. The sufficient condi-
tions that minimize the average total cost are the Hessian Matrix H > 0 at
tm, tm, tm, where
17.5 Discussion
Case 1: (0 < µ < ta < tb < tc < T)
Here, ν < ta and this implies that the demand curve becomes constant
before the production stops (ta = 7.2). Hence, the manufactured items will
take some time to sell and the inventory level will fall to zero after some
delay. This aspect is clear from the optimum values received in Example 1.
the level of the inventory reduces at a very slow rate. Hence, the level
reduces to zero after a long duration at tb = 8.05. This is clear from the opti-
mum values of tb, tc received in Example 2.
17.6 Inference
This study deals with a production-reserve (stock) model with a trapezoi-
dal-type demand for an item that perishes constantly and inventory that is
stored at a flexible cost. Shortages have been considered. The items consid-
ered are the ones like COVID gear (masks, PPE kits, etc.). Three cases have
been discussed related to the demand change time and the inventory levels
and production start and stop times. The decision variables are the pro-
duction halt/break off time (ta), the time at which the stock gets exhausted
(tb), and production restart time (tc). In the future, this study can be taken
further by considering a flexible deterioration rate.
References
1. E. Ritchie Practical inventory replenishment policies for a linear trend in
demand followed by a period of steady demand J. Oper. Res. Soc., 31 (7)
(1980), pp. 605-613.
2. An inventory model for deteriorating items with preservation facility of ramp
type demand and trade credit Ali Akbar Shaikh, Gobinda Chandra Panda,
Md. Al-Amin Khan, Abu Hashan Md. Mashud, Amiya Biswas https://doi.
org/10.1504/IJMOR.2020.110895 International Journal of Mathematics in
Operational Research Volume 17, Issue 4, 2020.
3. Open Journal of Business and Management, Vol.7 No.2, April 2019 An
Inventory Model for Ramp-Type Demand with Two-Level Trade Credit
Financing Linked to Order Quantity Hui-Ling Yang DOI: 10.4236/
ojbm.2019.72029
4. Chauhan et al. [2021] An order quantity scheme for ramp type demand
and back-logging during stock out with discount strategy, Anand Chauhan,
Mathematical Representation for Deteriorating Goods 263
18. Roy Chowdhury, R., Ghosh, S.K., Chaudhuri, K. S., (2014), ‘An Order-Level
Inventory Model for a Deteriorating Item with Time-Quadratic Demand
and Time-Dependent Partial Back-logging with Shortages in All Cycles’,
American Journal of Mathematical and Management Sciences, vol.33(2),
pp.75-97.
19. Tripathi, R.P., (2015), ‘Economic order quantity for deteriorating item with
nondecreasing demand and shortages under inflation and time discounting’,
International Journal of Engineering, Vol.28(9), pp.1295-1302.
18
An Amended Moth Flame Optimization
Algorithm Based on Fibonacci
Search Approach for Solving
Engineering Design Problems
Saroj Kumar Sahoo* and Apu Kumar Saha
Abstract
The moth flame optimization (MFO) algorithm is a swarm intelligence (SI) based
algorithm which gained popularity among researchers due to a special kind of
movement mechanism, namely, a transverse orientation mechanism of the moth
in nature. Like other SI based algorithms, it also suffers from good quality solution
and slow convergence speed. To avoid the drawbacks, a new variant of MFO algo-
rithm, namely, a Fibonacci technique based MFO algorithm (in short Ft-MFO)
is presented in this paper. We merged the concept of Fibonacci search method
in the classical MFO algorithm to improve the search quality and accelerate the
convergence speed of the MFO algorithm. To validate the performance of the pro-
posed algorithm, Ft-MFO is compared with six popular stochastic optimization
algorithms on an IEEE CEC2019 test suite and two constraint engineering design
problems. Experimental results demonstrate that the proposed Ft-MFO algorithm
is superior to the other stochastic algorithms in terms of solution quality and con-
vergence rate.
Keywords: Moth flame optimization algorithm, Fibonacci search method,
benchmark functions
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (265–282) © 2023 Scrivener Publishing LLC
265
266 Mathematics and Computer Science Volume 2
18.1 Introduction
The challenge of finding the better solution in optimization problems is an
interesting topic of research due to its importance in both academia and
industry. If the number of optimization parameters increases, the optimi-
zation problem complexity also increases. In recent decades, researchers are
very much interested in machine learning and artificial intelligence (AI) tech-
niques as real-life difficulties, such as constrained or unconstrained, linear or
nonlinear, continuous or discontinuous, that can be easily tackle by AI and
machine learning techniques [1, 2]. Due to the aforementioned character-
istics, there are various levels of difficulty to handle such types of difficulties
using conventional techniques through numerical or mathematical program-
ming, namely quasi-Newton approach, quadratic programming, conjugate
gradient, fast steepest method [3], etc. In various existing research [4], it has
been experimentally proved that the above-mentioned approaches are not
efficient enough to handle non-differentiable and real-life multimodal issues.
In contrast, the nature-based algorithms have played a substantial role in
tackling these issues, as these algorithms are simple and can be easily applied.
A few of them are the Genetic Algorithm (GA) [5], Particle Swarm Optimizer
(PSO) [6], Differential Evolution (DE) [7], Spotted Hyena Optimization
Algorithm (SHO) [8], Butterfly Optimization Algorithm (BOA) [9], Whale
Optimization Algorithm (WOA) [10], Sine Cosine Algorithm (SCA) [11],
Moth Flame Optimizer (MFO) [12], Salp Swarm Algorithm (SSA) [13], etc.
Typically, these algorithms initiate by a set of randomly chosen initial solu-
tions and continue until they get the optimal solution of a problem. When the
algorithm exceeds the pre-determined number of iterations it will automati-
cally be terminated. There is an increasing interest for the efficient, low-cost,
and effective implementation of such metaheuristic algorithms these days.
The MFO algorithm is the subject of this article. In 2015, Mirjalili dis-
covered MFO, a swarm intelligence-based algorithm. Transverse orienta-
tion is used by moths to navigate in the wild and served as an inspiration
for MFO. Spiral flight search and simple flame generation (SFG) are two
of the most important MFO tactics. In the SFG technique, the best moths
and flames collected thus far can be used to directly manufacture flames.
By following moths’ transverse direction, the SFS technique allows moths
to spiral toward the flames to update the positions in an iterative way. It is
possible that MFO will find the best option in the available search space.
The transverse orientation of moths in particular is critical to the effective-
ness of MFO. The key advantage of MFO on other algorithms is its abil-
ity to tackle several tough issues involving confined and unknown search
spaces such as optical network unit placement [14], automatic generation
An Amended Moth Flame Optimization Algorithm 267
MFO is formulated, namely Ft-MFO with the help of the Fibonacci technique.
The major steps involved in this work are as follows:
Moths move spirally when they are nearer to the flame, therefore the
author used a logarithmic spiral function which is as follows:
δ ⋅ e bt ⋅ cos(2π t ) + Fm (k ), i ≤ N .FM
K +1 i i
x i = bt (18.3)
δ i ⋅ e ⋅ cos(2π t ) + FmN . FM (k ), i ≥ N .FM
where = δ i xiK − Fmi represents the distance of the moth at ith place and
its specific flame (Fmi) and t can be any random number between −1 and 1.
Here, b is a fixed constant used to recognize the spiral flight shape. A moth
moves like a helix towards the flame with a one-dimensional approach and
a discrete value of t and represented as follows:
−1
a1 = −1 + current iter (18.4)
maxiter
t = (a1 – 1) × r + 1 (18.5)
where maxiter and a1 indicate the maximum iterations and the constant of
convergence, respectively, which decreases linearly from (−1) to (−2). In
every iteration, flame positions for the current and last iterations are col-
lected and arranged as per the fitness value for the global and local search.
The number of flames (N.FM) that have been lowered over the iteration
can be calculated using the formula below.
(N .FM Lst it − 1)
=N .FM round N .FM Lst it − crnt .it (18.6)
max it
270 Mathematics and Computer Science Volume 2
3
−iteration
r = −2 + e k ×maxiter
(18.7)
where k is a constant and its value is 0.55, which was suitably chosen so that
it helps in both global and local searches and is represented in Figure 18.1.
-1.1
MFO Ft-MFO
-1.2
-1.3
-1.4
Parameter (r)
-1.5
-1.6
-1.7
-1.8
-1.9
-2
0 100 200 300 400 500 600 700 800 900 1000
Number of iterations
F F
t1 LL + no − 2 (UL − LL), t 2 =
= UL − no − 2 (UL − LL) (18.9)
Fno Fno
End while
Output: The best solution with the minimum fitness
An Amended Moth Flame Optimization Algorithm 273
Table 18.1 Simulation results of Ft-MFO and other algorithms on CEC’2019 suite.
Algorithm F1 F2 F3 F4 F5
AVG STD AVG STD AVG STD AVG STD AVG STD
Ft-MFO 1 0 5 0 8.40 1.59 3.34e+01 3.39e+01 1.64 1.68
MFO 1 0 5 0 9.62 9.60e-01 1.05e+02 1.56e+01 1.00e+02 2.59e+01
DE 1.82E+06 8.25E+05 1.93e+03 3.40e+02 4.58 6.11e-01 8.54 1.99 1.77 2.64e-02
SOS 4.75E+03 6.88E+03 2.65e+02 1.96e+02 4.61 9.78e-01 2.08e+72 3.42e+72 1.19e+73 4.40e+73
JAYA 6.53E+06 2.83e+06 4.68e+03 7.78e+02 8.09 7.57e-01 4.10e+01 4.90 2.60 1.67e-01
BOA 1 0 5 0 6.79 1.15 1.26e+02 2.15e+01 1.09e+02 2.33e+01
WOA 1.45E+08 1.17E+08 1.10E+04 3.36E+03 8.16 1.60 6.62E+01 2.40E+01 3.83 1.43
Algorithm F6 F7 F8 F9 F10
AVG STD AVG STD AVG STD AVG STD AVG STD
Ft-MFO 3.07 1.21 8.22e+02 2.23e+02 4.25 2.41e-01 1.26 9.43e-02 2.15e+01 1.75e-01
(Continued)
An Amended Moth Flame Optimization Algorithm 275
Table 18.1 Simulation results of Ft-MFO and other algorithms on CEC’2019 suite. (Continued)
Algorithm F6 F7 F8 F9 F10
MFO 1.21e+01 1.09 2.29e+03 2.51e+02 5.13 1.79e-01 3.86 6.23e-01 2.16e+01 1.43e-01
DE 1.11 1.48e-01 1.07e+03 1.24e+02 4.48 2.71e-01 1.29 4.20e-02 2.25e+01 2.52
SOS 2.15 2.22e-01 4.55e+71 4.09e+71 5.50 0 1.59e+71 2.34e+71 2.18e+01 8.74e-04
JAYA 5.90 9.23e-01 1.51e+03 2.09e+02 4.28 2.08e-01 1.56 1.11e-01 2.16e+01 1.27e-01
BOA 1.36e+01 1.15 2.07e+03 2.89e+02 5.12 2.83e-01 4.14 5.51e-01 2.16e+01 1.85e-01
WOA 9.94 1.84 1.49E+03 2.85E+02 4.73 3.09E-01 1.48 1.69E-01 21.46 1.12E-01
276 Mathematics and Computer Science Volume 2
4 WOA
log(f(x)-f(x*))
MFO
MFO SOS
SOS 1.5
3
1
2
0.5
1
0
0 0 200 400 600 800 1000 1200 1400 1600 1800 2000
0 200 400 600 800 1000 1200 1400 1600 1800 2000 Function Evaluation
Function Evaluation
WOA WOA
MFO
log(f(x)-f(x*))
MFO
7 SOS 1.4 SOS
1.3
6.5 1.2
1.1
6 1
0.9
0 200 400 600 800 1000 1200 1400 1600 1800 2000
5.5 Function Evaluation
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Function Evaluation
1 2 3 D
D
A
4 A
results, using the global best solution in the optimization process quickly
brought the proposed method into focus. To avoid local optima traps and
premature convergence, the proposed technique is helpful.
In the future we can extend it to multi-objective optimization, apply it
to solve higher constraint optimization problems like car-side crash prob-
lems, robot gripper problems, welded beam design problems, etc. We can
generate an efficient metaheuristic algorithm by hybridizing our suggested
approach, Ft-MFO, with any other meta-heuristic algorithm.
References
1. Abbassi, R., Abbassi, A., Heidari, A. A., & Mirjalili, S. (2019). An efficient
salp swarm-inspired algorithm for parameters identification of photovoltaic
cell models. Energy Conversion and Management, 179, 362–372.
2. Faris, H., Ala’M, A.-Z., Heidari, A. A., Aljarah, I., Mafarja, M., Hassonah, M.
A., & Fujita, H. (2019). An intelligent system for spam detection and identi-
fication of the most relevant features based on evolutionary random weight
networks. Information Fusion, 48, 67–83.
3. McCarthy, J. F. (1989). Block-conjugate-gradient method. Physical Review
D, 40(6), 2149.
4. Wu, G., Pedrycz, W., Suganthan, P. N., & Mallipeddi, R. (2015). A variable
reduction strategy for evolutionary algorithms handling equality constraints.
Applied Soft Computing, 37, 774–786.
5. Holland, J. H. (1992). Genetic algorithms. Scientific American, 267(1),
66–73.
6. Kennedy, J., Eberhart, R. Particle swarm optimization. In: Proceedings of
ICNN’95—international conference on neural networks, Perth, Australia,
1995, 1942–1948.
7. Storn, R., & Price, K. (1997). Differential evolution–a simple and efficient
heuristic for global optimization over continuous spaces. Journal of Global
Optimization, 11(4), 341–359.
8. Dhiman, G., & Kumar, V. (2017). Spotted hyena optimizer: A novel bio-
inspired based metaheuristic technique for engineering applications.
Advances in Engineering Software, 114, 48–70.
9. Arora, S., & Singh, S. (2015). Butterfly algorithm with levy flights for
global optimization. 2015 International Conference on Signal Processing,
Computing and Control (ISPCC), 220–224.
10. Mirjalili, S., & Lewis, A. (2016). The whale optimization algorithm. Advances
in Engineering Software, 95, 51–67.
11. Mirjalili, S. (2016). SCA: A sine cosine algorithm for solving optimization
problems. Knowledge-Based Systems, 96, 120–133.
280 Mathematics and Computer Science Volume 2
26. Ma, L., Wang, C., Xie, N., Shi, M., Ye, Y., & Wang, L. (2021). Moth-flame
optimization algorithm based on diversity and mutation strategy. Applied
Intelligence.
27. Sahoo, S. K., Saha, A. K., Sharma, S., Mirjalili, S., & Chakraborty, S. (2022).
An enhanced moth flame optimization with mutualism scheme for function
optimization. Soft Computing, 26(6), 2855–2882.
28. Sahoo, S. K., & Saha, A. K. (2022). A hybrid moth flame optimization
algorithm for global optimization. Journal of Bionic Engineering, 19(5),
1522–1543.
29. Sahoo, S. K., Saha, A. K., Nama, S., & Masdari, M. (2022). An improved moth
flame optimization algorithm based on modified dynamic opposite learning
strategy. Artificial Intelligence Review, 1–59.
30. Sahoo, S. K., Saha, A. K., Sharma, S., Mirjalili, S., & Chakraborty, S. (2022).
An enhanced moth flame optimization with mutualism scheme for function
optimization. Soft Computing, 26(6), 2855–2882.
31. Sahoo, S.K., Saha, A.K., Ezugwu, A.E. et al. Moth flame optimization: theory,
modifications, hybridizations, and applications. Arch Computat Methods
Eng (2022). https://doi.org/10.1007/s11831-022-09801-z.
32. Chakraborty, S., Saha, A. K., Sharma, S., Sahoo, S. K., & Pal, G. (2022).
Comparative performance analysis of differential evolution variants on engi-
neering design problems. Journal of Bionic Engineering, 19(4), 1140–1160.
https://doi.org/10.1007/s42235-022-00190-4.
33 Sahoo, S. K., Sharma, S., & Saha, A. K. (2023). A Novel variant of moth flame
optimizer for higher dimensional optimization problems, 1–27.
34 Sahoo, S. K., & Saha, A. K. (2022, August). A modernized moth flame optimi-
zation algorithm for higher dimensional problems. In ICSET: International
Conference on Sustainable Engineering and Technology, (Vol. 1, No. 1, pp.
9–20).
35. Chakraborty, S., Kumar Saha, A., Sharma, S., Mirjalili, S., & Chakraborty, R.
(2021). A novel enhanced whale optimization algorithm for global optimiza-
tion. Computers & Industrial Engineering, 153, 107086.
36. Nama, S., Saha, A. K., & Ghosh, S. (2017). A hybrid symbiosis organisms
search algorithm and its application to real world problems. Memetic
Computing, 9(3), 261–280.
37. Sharma, S., & Saha, A. K. (2020). m-MBOA: A novel butterfly optimiza-
tion algorithm enhanced with mutualism scheme. Soft Computing, 24(7),
4809–4827.
38. Chakraborty, S., Saha, A. K., Chakraborty, R., Saha, M., & Nama, S. (2022).
HSWOA: An ensemble of hunger games search and whale optimization algo-
rithm for global optimization. International Journal of Intelligent Systems,
37(1), 52–104.
39. Sharma, S., Chakraborty, S., Saha, A. K., Nama, S., & Sahoo, S. K. (2022).
mLBOA: A Modified Butterfly Optimization Algorithm with Lagrange
Interpolation for Global Optimization. Journal of Bionic Engineering.
282 Mathematics and Computer Science Volume 2
Abstract
Medical image segmentation consists of heterogeneous pixel intensities, noisy/
ill-defined borders, and high variability, which are significant technical obstacles
for segmentation. Also, generally the requirement of annotated samples by the
networks is significantly large to achieve high accuracy. Gathering this dataset for
the particular application and annotating new images is both time-consuming and
costly. Unet solves this problem by not requiring vast datasets for picture segmen-
tation. The present work describes the use of a network that depends on augmen-
tation of the existing annotated dataset to make better use of these examples and
a comparison of encoder accuracy on Unet is presented. The encoder principal
function is to reduce image dimensionality while keeping as much information
as possible. EfficientNets tackles both of these issues and utilizing it as an encoder
of Unet can further enhance its accuracy. The test dataset highest F1-Score and
IoU were 0.7655 and 0.6201 on neuronal data values, respectively. It outperforms
Inception and ResNet encoder networks with considerably more parameters and
a higher inference time.
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (283–290) © 2023 Scrivener Publishing LLC
283
284 Mathematics and Computer Science Volume 2
19.1 Introduction
Neurodegenerative illnesses cover a diverse spectrum of ailments result-
ing from progressive damage to cells and nervous system connections.
These illnesses eventually lead to diseases such as Alzheimer’s and
Parkinson’s disease [1], which are among the six highest causes of death.
According to the WHO report [2] on neurological disorders, neuro-
logical illnesses account for 12% of all fatalities worldwide. Figure 19.1
shows the architecture of the modified U-Net model used for training.
The goal of finding treatments and solutions for neurodegenerative ill-
nesses is becoming more urgent with the increase in population. In 2015,
neurological illnesses caused 250.7 million disability-adjusted life years
(DALYs) and 9.4 million deaths, accounting for 10.2% of global DALYs
and 16.8% of global deaths [3]. Our work is focused on assisting doctors
in determining the effects of medical treatments in neuron cells using
machine learning and computer vision.
For biomedical image analysis, the segmentation of 2D pictures is a
critical problem [4]. We have proposed an Unet based neural network
that upgrades the FCN model for biomedical application, i.e., RIC
Unet, for nuclei segmentation. Following Unet architecture, RIC-Unet
uses a channel attention mechanism, residual blocks, and multi-scale
ENCODER DECODER
I O
N U
P T
U P
T (16x16x2048)
U
(8x8x1824)
(16x16x3104)
T
(16x16x768) (16x16x1056) (32x32x640)
(32x32x256) (64x64x128) (64x64x368)
(32x32x384)
(64x64x240) (128x128x64) (128x128x208)
(256x256x32)
(128x128x144)
(128x128x48)
Concatenate
Figure 19.1 Architecture of Unet with EfficentNet as encoder. The decoder is similar to the
original Unet model but with cropping and contamination from various encoder elements.
Image Segmentation of Neuronal Cell with Unet Architecture 285
19.2 Methods
The Unet framework mainly inspired our architecture. We utilized
EfficientNet-B0 as an encoding part. Like Unet, encoder branch and
decoder branch features are concatenated. We chose EfficientNet as the
encoder because it proposes an effective compound scaling strategy that
effectively scales up a baseline ConvNet to any target resource limitations
in a more principled manner while retaining model efficiency.
The ImageNet dataset containing color images was used for training
the original EfficientNet. The images had a resolution of 224×224. In the
present study, a pre-trained EfficientNet model was implemented as an
encoder part of Unet to reduce the training time. The EfficientNet-B0 with
low parameter quantity was applied owing to limited complexity and an
easily identifiable nature of the neuronal images.
19.3 Dataset
The dataset consists of 606 grayscale images with an image size of 520 ×
704 pixels. We chose 364 photos at random for training, 61 for validation,
and 60 for testing. Image count of ‘shsy5y’, ‘Astro’, ‘Cort’ was 155, 131, and
286 Mathematics and Computer Science Volume 2
320, respectively. The neuroblastoma cell line Cort consistently has the
lowest precision scores of the three cancer cell lines. This could be because
neuronal cells have a distinct, uneven, and concave look, making them dif-
ficult to segment using standard mask heads.
1 N
AIOUαmask= ∑i =1 IOUαmask (19.1)
N
19.6 Result
For this work of semantic segmentation, we used a variety of designs
and the results are given in Table 19.1. Our investigation began with the
DeepLabV3+ framework. For extraction of features, we initially utilized
EfficientNet-B0 as an encoder. We experimented with several encoder back-
bones such as EfficientNet-B0 [8], ResNet-50 [9], InceptionResNetV2 [10],
Image Segmentation of Neuronal Cell with Unet Architecture 287
Table 19.1 Result comparison in terms of mean intersection over union (IoU),
F1-score, and accuracy on the neuronal dataset.
Model F1-score IoU Accuracy
Unet with EfficientNet-B0 as encoder 0.7286 0.573 0.9466
(binary cross entropy logits)
Unet with ResNet-50 as encoder (binary 0.7053 0.5447 0.9254
cross entropy logits)
Unet with Inception-ResNet-v2 as 0.7264 0.5704 0.9550
encoder (binary cross entropy logits)
Unet with EfficientNet-B0 as encoder 0.7655 0.6201 0.9493
(dice loss)
PSPNet with EfficientNet-B2 as encoder 0.6075 0.4362 0.9321
(binary cross entropy logits)
PSPNet with InceptionNet as encoder 0.6579 0.4902 0.9311
(binary cross entropy logits)
DeeplabV3+ with EfficientNet-B5 as 0.7089 0.5490 0.9351
encoder (binary cross entropy logits)
Unet with Inception-ResNet-v2 as 0.7283 0.5727 0.9166
encoder (dice loss)
DeeplabV3+ with EfficientNet-B5 as 0.7265 0.5705 0.9351
encoder (dice loss)
and EfficientNet-B5 with the renowned Unet decoder. From Table 19.1,
it can be shown that the Unet decoder outperforms the DeepLabV3+
decoder. That is because Unet now has additional low-level encoder char-
acteristics that are essential in analyzing complicated scenes with multiple
dense objects.
We further analyze the class-wise segmentation performance of the
best-performing model on the three classes. We obtain class-wise IoU on
the validation dataset, shown in Table 19.2. It can be observed that the Cort
Table 19.2 Class wise IoU on neuronal dataset with efficient Unet as model.
Neuronal cells shsy5y Astro Cort
Mean IoU 0.4580 0.5474 0.2948
288 Mathematics and Computer Science Volume 2
cell has the lowest IoU. The Cort cell image contains different geometry
types compared to Astro and sphy5y. Besides the circular geometry of Cort
cells, thread-like structures are also present in the image, which is similar
to Astro and shyishy5 geometry. This makes model prediction clouded.
The model’s accuracy is further hampered by a small dataset. The ground
truth segmentation maps of a few images, as well as their anticipated seg-
mentation maps, are shown in Figure 19.2.
The first column displays the input photos for a single cell. The ground
truth and anticipated segmentation maps are shown in the second and
third columns, respectively. The cell type is in the order Astro, Cort,
Shsy5y from top to bottom.
Figure 19.2 Result of semantic segmentation on neuronal validation dataset with efficient
Unet as architecture.
Image Segmentation of Neuronal Cell with Unet Architecture 289
19.7 Conclusion
This study employs multiple kinds of encoders for image segmentation
and is based on Unet architecture. It provides a comprehensive procedure
that includes selecting a data set, acquiring a training set, training a deep
convolutional neural network, and segmenting cell images using the con-
volution neural network. Finally, following cell image segmentation, the
Unet with various encoders was utilized to generate the resultant image,
and a basic analysis of the acquired values was performed. This technique
achieves high-precision semantic segmentation of shsy5y cells and Astro
pictures compared to traditional image segmentation.
References
1. S. Przedborski, M. Vila, and V. Jackson-Lewis, ‘Series Introduction:
Neurodegeneration: What is it and where are we?’, J. Clin. Invest., vol. 111,
no. 1, p. 3, Jan. 2003, doi: 10.1172/JCI17522.
2. W. H. Organization, ‘Neurological disorders: public health challenges’, World
Health Organization, 2006.
3. V. L. Feigin et al., ‘Global, regional, and national burden of neurological dis-
orders, 1990–2016: a systematic analysis for the Global Burden of Disease
Study 2016’, Lancet Neurol., vol. 18, no. 5, pp. 459–480, May 2019, doi:
10.1016/S1474-4422(18)30499-X.
4. Z. Zeng, W. Xie, Y. Zhang, and Y. Lu, ‘RIC-Unet: An Improved Neural
Network Based on Unet for Nuclei Segmentation in Histology Images’,
IEEE Access, vol. 7, pp. 21420–21428, 2019, doi: 10.1109/ACCESS.2019.
2896920.
5. J. Zhuang, ‘LadderNet: Multi-path networks based on U-Net for medical
image segmentation’, Oct. 2018, Accessed: Dec. 13, 2021. [Online]. Available:
https://arxiv.org/abs/1810.07810v4.
6. C. A. R. Goyzueta, J. E. C. De la Cruz, and W. A. M. Machaca, ‘Integration
of U-Net, ResU-Net and DeepLab Architectures with Intersection Over
Union metric for Cells Nuclei Image Segmentation’, pp. 1–4, Nov. 2021, doi:
10.1109/EIRCON52903.2021.9613150.
7. M. Everingham, L. Van Gool, C. K. I. Williams, J. Winn, and A. Zisserman,
‘The pascal visual object classes (VOC) challenge’, Int. J. Comput. Vis., vol. 88,
no. 2, pp. 303–338, Sep. 2010, doi: 10.1007/s11263-009-0275-4.
8. M. Tan and Q. V. Le, ‘EfficientNet: Rethinking Model Scaling for Convolutional
Neural Networks’, 36th Int. Conf. Mach. Learn. ICML 2019, vol. 2019-June,
pp. 10691–10700, May 2019, Accessed: Dec. 13, 2021. [Online]. Available:
https://arxiv.org/abs/1905.11946v5.
290 Mathematics and Computer Science Volume 2
9. K. He, X. Zhang, S. Ren, and J. Sun, ‘Deep Residual Learning for Image
Recognition’, Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit.,
vol. 2016-December, pp. 770–778, Dec. 2015, doi: 10.1109/CVPR.2016.90.
10. C. Szegedy, S. Ioffe, V. Vanhoucke, and A. A. Alemi, ‘Inception-v4, Inception-
ResNet and the Impact of Residual Connections on Learning’, 31st AAAI
Conf. Artif. Intell. AAAI 2017, pp. 4278–4284, Feb. 2016, Accessed: Dec. 15,
2021. [Online]. Available: https://arxiv.org/abs/1602.07261v2.
20
Automorphisms of Some
Non-Abelian p−Groups of Order p4
Muniya Sansanwal1*, Harsha Arora2 and Mahender Singh3
Department of Mathematics, JJTU Jhunjhunu, Rajasthan, India
1
3
Department of Mathematics, Om Sterling Global University, Hisar, Haryana, India
Abstract
This paper contains the number of automorphisms of several non-Abelian groups
of order p4, p-odd prime are computed and GAP (Group Algorithm Programming)
software has been used for the verification of a number of automorphisms.
20.1 Introduction
Let G be a p-group of order p4, p-odd prime and Aut(G) represent the group
of all automorphisms of a given group G. There are many research papers
in the literature related to the automorphisms of p-groups, for instance [1,
3, 4], etc. In [1], the automorphisms of groups of order p3 are computed
along with the automorphisms of abelian groups of order p4. The present
paper is an extension of the research work in [1]. In this paper, we will
compute the automorphisms of groups of order p4. William Burnside [2]
classified all groups having order p4. By using these classifications we derive
the automorphisms of some groups having order p4. We divide the derived
results into two parts. In these two parts, one part is dedicated to the cat-
egorization of all p−groups of order p4 and the other part is dedicated to
investigating the number of automorphisms of several non-Abelian groups
with order p4.
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (291–304) © 2023 Scrivener Publishing LLC
291
292 Mathematics and Computer Science Volume 2
Non-Abelian Groups:
3 2
p
G6 =< x1 , x2 : x1= p
x=
2 , x2 x1 x11+ p x2 > ≅ Z p3 φ Z p ,
1=
φ ( y ) ↔ (1 + p2 ) y .
2 2
p
G7 =< x1 , x2 : x=
1
p
x=
2 , x2 x1 x11+ p x2 > ≅ Z p2 φ Z p2 ,
1=
φ ( y ) ↔ (1 + p) y .
2
p p
G8 =< x1 , x2 , x3 : x=
1 x=2 = x3p 1= , x1x2 x2 x1 , x2 x3 = x3 x2 ,
x3 x1 = x1x2 x3 > ≅ (Z p × Z p ) Z p2 .
p p p p
G9 =< x1 , x2 , x3 , x 4 : x=
1 x=2 x=3 x=
4 1, x 4 x3 = x3 x 4 , x2 x 4 = x 4 x2 ,
= x1x 4 x= 4 x1 , x 2 x 3 x3 x2 , x2 x1x3 = x3 x1 ,
x=1x 2 x2 x1 > ≅ ((Z Z p × Z p ) Z p ) × Z p .
p p p p
For pp >> 33,,G
G1010 =
<< xx1 ,,xx2 ,,xx3 ,,xx4 :: xx=
= 1 2 3 4 =
1p
1
xx=
2p
=2
xx=
3p
=3
x==
x=
=4 1,, xx44xx33 xx=
4p 1 2x
= 3 x 4 , x 4 x2
2 x3 x 4 , x 4 x2
xx1xx2 xx4 ,,
1 2 4
x3 x= 2 x3 x 4 , x 4 x2 x1x= 2 x 4 , x1x 4
= x1x 4 xx=
4x
= 1 , x2 x3
4 x1 , x 2 x 3
xx=
3x
= 2 , x1x 3
3 x 2 ,2 x1x 3
xx=
3x
= 1 , x1x 2
3 x1 , x1x 2
xx2 xx1 >>≅≅ ((ZZp ×× ZZp ×× ZZp )Z
2 1 p p p )Z
ZZp ,,
p
−p
> ≅ (Z p × Z p × Z p )ZZ p ,If p = 3 then < x1 , x2 x3 : x1p= x= 2
p
x =p
3 = 1, x x
1 2 x= x ,
2 1 1 3x x x 3 1 2 x 2 x 3 = x 3 x1 x 2 > .
x x ,
x=
2 x1 , x1x 3 x3 x1x2 , x2 x3 = x3 x1− p x2 > .
Automorphisms of Some Non-Abelian p−Groups of Order p4 293
2
1+ p
G11 = < x1 , x2 , x3 : x1p= x= p
2 x=
=
3
p
1, x2 x1 x= 1 x 2 , x1x 3 x3 x1 ,
x 2 x 3 = x 3 x 2 > ≅ ( Z p2 Z p ) × Z p .
2
G12 = < x1p= x=
p
2
p
x=
3=1, x2 x1 x=
1x 2 , x1x 3 x3 x1 ,
x3 x2 x1p x2 x3 > ≅ (Z p2 × Z p ) Z p .
=
2
1+ p
G13 = < x1 , x2 , x3 : x1p= x=
p
2
p
x=
=
3 1, x2 x1 x=
1 x 2 , x 3 x1 x1x2 x3 ,
x=
2 x3 x3 x2 > ≅ (Z p2 × Z p )φ1 Z p , φ1(z ) ↔ (1,1, 0)z .
2
1+ p
For p > 3,G14 = p
< x1 , x2 , x3 : x1= p
x=
2
p
x=
=
3 1, x2 x1 x=
1 x 2 , x 3 x1 x11+ p
, x3 x2 x1p x2 x3 > ≅ (Z p2 Z p )φ2 Z p , φ2 (z ) ↔ (1,1,1)z ,
x2 x3=
2
If p = 3 then < x1 , x2 , x3 : x1p = x2p =
1, x3p x=
= p
1 , x1x 2 x2 x11+ p ,
x1x3 = x3 x1x2−1 , x=
3 x2 x2 x3 > .
2
1+ p
Forp p> >3,3G,G
15 = < <x1x, x,2x, x,3x: x: 1xp=p=2
x= p
p x= = p
1, x2 x1 x=
p 1, x 2 x1 1+xp2 , x 3 x1 x11x+1dp+dp
x2xx3x, ,x3xx2x
15 =
2x= 3x== 1x=
1 2 3 1 2 3 1 x 2 , x 3 x1 1 2 3 3 2
1+ p 1+dp
x1 x2 , x3 x1 x1 x2 x3= , x3 x2 x1 dpdp
x x > ≅ ( Z Z ) Z , φ ( z ) ↔ (1 ,1 , d ) z
where
z d ≢
= x1 2x22x2 > ≅ (Z p 2 Z φ3 Z p , φ3 (z ) ↔ (1,1,d )
2 p )φ 3 p 3
p
1,d ) z 0,1(mod p), 2
p
−p
If p= 3,< x1 , x2 , x3 ; x1p= x= p
2= 1, x3p x= 1 , x1 x 2 x2 x11+ p , x1x3 = x3 x1x2−1 ,x= 3 x2 x2 x3 > .
1+ p −1
1 , x1x 3 = x 3 x1x 2 ,x=3 x2 x2 x3 > .
⇒ x1(m+i +s ) x2(n+qi + j+qs +ks +t ) x3(q+k +u ) x 4(r +l +v ) = x1( s +i ) x2(t +ui + j ) x3(u+k ) x 4(v +l )
⇒ x1m x2(n+qi +qs +ks ) x3q x 4r = x2(ui )
⇒ m ≡ 0(mod p),
n + qi + qs + ks ≡ ui(mod p),
q ≡ 0(mod p) and
r ≡ 0(mod p)
⇒ n + ks ≡ ui(mod p)
or n ≡ iu − sk(mod p)
Automorphisms of Some Non-Abelian p−Groups of Order p4 295
⇒ n = iu - sk ≢ 0(mod p).
h1(h1 − 1) h (h − 1)
jh1 + ik + nh2 + th3 + 3 3 su + xh4 + kh1sh3 ≡ 0(mod p),
2 2
(20.2)
kh1 + uh3 ≡ 0(mod p), (20.3)
lh1 + vh3 + zh4 ≡ 0(mod p). (20.4)
Equations (20.1) and (20.3) are the system of two homogeneous equa-
tions with iu − sk ≠ 0 determinant, which gives only a trivial solution.
Hence, h1 = h3 = 0. From (20.4), we get
zh4 ≡ 0(mod p)
As
z ≢ 0(mod p),
we get
h4 = 0.
Further, from (20.2) we get nh2 ≡ 0(mod p), but n ≢ 0(mod p) ⇒ h2 = 0.
( )
So, if h ∈ ( (Z p × Z p ) Z p ) × Z p and ψ(h) = 1, it must be that h = 1.
Hence, the kernel is trivial and ψ is an automorphism with the con-
straints on the parameters already deduced. Now it is very easy to calculate
the order of the automorphism group. We have p choices for all j, l, t, v, x,
and p − 1 choices for z and i, u, s, and k will in some sense be equivalent
to a matrix in GL(Fp), which gives (p2 −p)(p2 − 1) choices for the elements.
( )
Therefore, we have that ψ ∈ Aut ( (Z p × Z p ) Z p ) × Z p is really defined as:
Aut ( ( (Z p × Z p ) Z p ) × Z=
p) p6 ( p − 1)2 ( p2 − 1).
Automorphisms of Some Non-Abelian p−Groups of Order p4 297
By using our relations from the structure of G10 and elementary calcula-
tions, we can deduce some useful relations:
ψ(x4)ψ(x3) = ψ(x2)ψ(x3)ψ(x4)
ψ (x 4 )ψ (x3 ) = ψ (x2 )ψ (x3 )ψ (x 4 )
w x y z s t u v
x1m x2n x3q x 4r x1s x2t x3u x 4v x1w x2x x3y x 4z
⇒x x x x x x x x =
1 2 3 4 1 2 3 4
zu ( z −1)
( zt + )
( zu )
⇒ x1 2 x 2
ru (r −1) vy (v −1) ry (r −1)
(m+rt +vx + + +rx +rvy + )
(n+ru +vy +ry ) q r
=x 1
2 2 x x x 2
2 3 4
r ≡ 0 (mod p) (20.8)
zu(z − 1) vy(v − 1)
⇒ zt + ≡ m + vx + (mod p), (20.9)
2 2
zu ≡ n + vy(mod p) (20.10)
Also,
x 4 x2 = x1x2 x 4
⇒ ψ (x 4 )ψ (x2 ) = ψ (x1 )ψ (x2 )ψ (x 4 )
⇒ x1w x2x x3y x 4z x1m x2n x3q x 4r = x1i x2j x3k x 4l x1m x2n x3q x 4r x1w x2x x3y x 4z
ly (l −1)
i +m+w +ln+lx +
(w +m+ zn) ( x +n) y z 2 ( j +n+ x +ly ) ( k + y ) (l + z )
⇒x 1 x
2 x x =x
3 4
1 x 2 x
3 x4
ly (l −1)
i +ln+lx +
2 ( j +ly ) k l
⇒ x1( zn) =
x
1 x 2 x x
3 4
ly(l − 1)
⇒ zn ≡ i + ln + lx + (mod p), (20.11)
2
l ≡ 0(mod p) (20.14)
⇒ ψ (x2 )ψ (x3 ) =
ψ (x3 )ψ (x2 )
⇒ x1m x2n x3q x 4r x1s x2t x3u x 4v =
x1s x2t x3u x 4v x1m x2n x3q x 4r
⇒ x1(m+s ) x2(n+t ) x3u x 4v =
x1( s +m+vn) x2(t +n) x3u x 4v
⇒ vn ≡ 0(mod p).
I ≡ z2u(mod p).
h (h −1)
(nh2 +th3 + xh4 + 4 4 yz )
(uh3 + yh4 ) (zzh4 )
x 2
2 x 3 x4 =1
h4 (h4 − 1) h (h − 1)z (z − 1)
⇒ ih1 + mh2 + sh3 + wh4 + xz + 4 4
2 4
h (h − 1)(h4 − 2) 2
y+ 4 4 yz ≡ 0(mod p), (20.17)
6
h4 (h4 − 1)
nh2 + th3 + xh4 + yz ≡ 0(mod p), (20.18)
2
300 Mathematics and Computer Science Volume 2
Aut ( (Z p × Z p × Z p ) Z=
p) p5 ( p − 1)2 .
x1pi ≠ 1
⇒ p i.
Next, we find the constraints for parameters using the kernel. Let
x1x x2y x3z be any element of (Z p2 Z p ) × Z p that it is mapped to identity.
x(x − 1)
p2 /xi + ijp + pty + puz
2
x(x − 1)
⇒ p/p2 /xi + ijp + pty + puz
2
x(x − 1)
⇒ p/xi + ijp + pty + puz
2
⇒ p/xi
But p i
⇒ p/x
⇒x= 0(mod p).
my ≡ 0(mod p)
But p f m
⇒ y ≡ 0(mod p).
p6 ( p − 1)2 .
|Aut (Z p2 Z p ) × Z p | =
References
1. H. Arora and R. Karan: What is the probability an automorphism fixes a group
element?, Communications in Algebra, 45(3), 1141–1150 (2017).
2. William Burnside: Theory of groups of finite order, Cambridge University
Press, first edition, 1897. Reprinted 2010 through Nabu Press.
3. Geir T. Helleloid Automorphism Groups of Finite p-Groups: Structure and
Applications, arxiv: 0711.2816 (2007).
4. Hans Liebeck: The Automorphism Group of finite p-groups, Journal of Algebra,
4, 426–432 (1966).
21
Viscoelastic Equation of
p-Laplacian Hyperbolic Type with
Logarithmic Source Term
Nazlı Irkıl* and Erhan Pişkin†
Department of Mathematics, Dicle University, Diyarbakır, Turkey
Abstract
This paper aims to address the problem with viscolelastic p-Laplacian type equa-
tions with logarithmic nonlinearity,
t
u tt +
∫
0
g(t − s)∆ u(s) ds − div((1 + ∇ p − 2 )∇u) − ∆u t = |u|p − 2 uln|u|
21.1 Introduction
Let Ω be a bounded domain in Rn (for n ≥ 1) with smooth boundary Σ =
∂Ω. We deal with the following problem in the initial boundary problem
for (x, t) ∈ Ω × R+:
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (305–326) © 2023 Scrivener Publishing LLC
305
306 Mathematics and Computer Science Volume 2
t
tt
0
∫
u + g(t − s) ∆ u(s) ds − div((1 + ∇( p − 2) )∇u) − ∆u = |u|( p − 2)uln|u|,
t
u(x,0) = u 0 (x), u t (x , 0) = u1 (x),
u(x , t) = 0, x ∈∑ × R + ,
The memory kernel g(t) is a real function which has typical properties.
In (21.1), if we take ln |u| = 1 and replace the second order as the fourth
order, we have
t
2
∫
utt + ∆ u + g(t − s) ∆ u(s) ds − div(∇ p − 2∇u) − ∆u t = f (u).
0
Equation (21.2) was named as a fourth order weak viscoelastic plate equa-
tion with a lower order perturbation of p-Laplacian type and it can be taken
as a general form of the one-dimensional nonlinear equation of elastoplastic
microstructure flows. This type of equation was first considered by Andrade
et al. [3]. They deal with interplay of the p-Laplacian operator with viscoelastic
terms. The authors obtained existence and decay results of solutions with gt(t)
≤ cg (t) . Later results about qualitative theory results of solutions to (21.2) type
equations were obtained by different authors (see [4, 13, 14, 18, 19, 26]).
The logarithmic wave equations which were introduced by Bialynicki-
Birula and Mycielski in [5] are related with many areas of physics.
Logarithmic source term occurs naturally in inflation cosmology and
quantum mechanics, supersymmetric field theories, and nuclear physics
[7, 17]. Later, by the motivation of these works, analysis of logarithmic
source terms attracted many authors (see [8−10, 16, 23, 25, 32, 35]). There
is some work about mathematical behavior in viscoelasticity with logarith-
mic source terms. The references were given in our work for interested
readers (see [2, 6, 11, 12, 20, 31]).
The model (21.1) for g = 0 and without ∆u was studied, but there is
not a substantial number of work related with p-Laplacian hyperbolic type
equations. For contact problems on p-Laplacian type equations with loga-
rithmic source terms see ([15, 24, 30, 33, 34]).
The goal of our work is to prove the global existence of a weak solution
to the problem (21.1). Blow up results were established for suitable condi-
tions at E (0) < d. As far as know, there is no work which is related on p-
Laplacian hyperbolic type equations with viscoelastic terms and logarith-
mic source terms.
Viscoelastic Equation of p-Laplacian Hyperbolic Type 307
21.2 Preliminaries
In this part, we consider the notation and some well-known lemmas. We
define the inner product as
(v , u) =
∫ v(x )u(x )dx.
in Standard Lebesque Space L2 (Ω) and we denote:
(∇v, ∇u) = ∫ ∇v(x)∇u(x)dx
in H 01 (Ω ). The norm in L2 (Ω) and Lp (Ω) will be defined by ǁ.ǁ and ǁ.ǁ p. And
C will be various positive constants.
Lemma 21.1. [1, 22] Let q be a positive number, which implies that
2 ≤ q < ∞, for n = 1,2
2n
2≤q≤ , for n ≥ 3.
n−2
So, for ∀ u ∈ H 01 (Ω ), we have
g (0) ≥ 0 ,
∫ g (s)ds = 1 − l > 0,
0
(21.4)
E(t) will be denoted total energy functional for the model investigated
in (21.1). It follows that
t
1 1 1 1
E(t ) = ||ut ||2 + ||∇ u||pp + 1 −
2 p 2 ∫0
g (s )ds ||∇ u||2 + 2 ||u||pp
p
1 1
+ ( g °∇ u)(t ) −
2 p ∫ u ln|u| dx
Ω
p (21.6)
t
1 1 1
p
p
J (u) = ||∇ u||p + 1 −
2 ∫0
g (s )ds ||∇ u||2 + 2 ||u||pp
p
1 1
+ ( g °∇ u)(t ) −
2 p ∫ u ln|u| dx.
Ω
p
(21.7)
t
p
I (u) = ||∇ u|| + 1 −
p
∫
0
g (s )ds ||∇ u||2 + ( g °∇ u)(t ) −
∫ u ln|u| dx ,
Ω
p
(21.8)
where
t
( g °∇ u)(t ) = ∫ 0 g (t − s )||∇ u(s ) −∇ u(t )||2 ds (21.9)
t
1 1 I (u) 1
J (u) = − 1 −
2 p ∫
0
g (s ) ||∇ u||2 +
p
+ 2 ||u||pp
p
1 1
+ − ( g °∇ u)(t ), (21.10)
2 p
Viscoelastic Equation of p-Laplacian Hyperbolic Type 309
1
E(t ) = ||ut ||2 + J (u). (21.11)
2
According to potential well theory, the set stability for the problem
(21.1) was considered as:
Now, we state some definitions and lemmas. Lemma 21.5 and Lemma
21.6 are related with potential well theory results.
( p − 2)a 2 a2 n
∫
Ω
u p ln |u|dx <
4π
||u||pp +
2π
||∇ u||pp − (1 + ln a)||u||pp + ||u||pp ln ||u||p
p
and
t
∫Ω
utt vdx +
∫Ω
(1 +∇ p−2 )∇ u∇ vdx −
∫ ∫ ∫ ∇ u∇ vdsdx + ∫ ∇ u∇ vdx = ∫ u
Ω 0 Ω Ω
t
Ω
p−2
uln |u|vdx,
for each test functions v ∈W01,p (Ω ), then u will be taken as a weak solution
for the (21.1).
Lemma 21.4. Assume that (A1) and (A2) conditions, which are related with
g(t) functional, hold. Then, the total energy of problem (21.1) decreases
over respect to t because
1 1 g (t )
E ′(t ) = ( g ′ °∇ u)(t ) − ||∇ ut ||2 − ||∇ u(t )||2
2 2 2
1 (21.15)
≤ − ||∇ ut ||2 ≤ 0.
2
where
( g ′ °∇ u)(t ) =
∫ g ′(t − s)||∇ u(s) −∇ u(t )|| dt
0
2
(21.16)
t t
1 1
E(t ) = E(0) − ||∇ ut ||2 −
2 ∫ 0
2
g (t )||∇ u(t )|| +
2 ∫ ( g ′ °∇ u)(t )
0
1
≤ ||∇ ut ||2 (21.17)
2
dJ (λu)
=0
d λ λ = λ*
iii) J (λu) functional is strictly decreases on λ* < λ < ∞, increases on
0 < λ < λ* and attains the maximum at λ = λ* and I (λu) > 0 for
0 < λ < λ*, I (λu) < 0 for λ* < λ < ∞ and the Nehari functional
dJ (λu)
I(λ*u) = 0, where λ = I (λu).
dλ
Proof. i) Thanks to the definition of J(u), we conclude that
t
1 1
J ( λu ) =
p ∫
Ω
p
(λ∇ u) dx + 1 −
2 ∫
0
g (s )ds
∫ (λ∇ u) dx
Ω
2
1 1 1
−
p ∫ (λu) ln |λu|dx + p ∫ (λu) dx + 2 λ ( g °∇ (λu))(t )
Ω
p
2
Ω
p 2
t
λ2 λ2 1
∫
p
= 1 − g (s )ds || u||2 + ( g °∇ u)(t ) + 2 λ p ||u||p
2 2 p
0
p
λ λp 1 p
∫ ∫ (u) dx.
p
+ ||∇ u||p − (u) p ln|u| dx − λ lnλ p
(21.18)
p p p
Ω Ω
p
It is easy to get from ||u||p ≠ 0 that is lim J (λu) = 0 , lim J (λu) = −∞.
λ →0 λ →∞
ii) Now, differentiating J (λu) with respect to λ, it yields that
t
dJ (λu)
dλ
= λ 1 −
∫
0
g (s )ds || u||2 +( g °∇ u)(t ) + λ p − 2 ||∇ u||pp
−λ p−2
∫ (u) ln|u|dx − λ
Ω
p p−2
lnλ||∇ u||pp .
(21.19)
dJ (λu)
Let λ −1 = φ (λu), so we obtain
dλ
312 Mathematics and Computer Science Volume 2
dφ (λu)
dλ
= − λ p −3[(2 − p)||∇ u||pp +( p − 2)
∫ (u) ln|u|dx
Ω
p
p p
+ ( p − 2)lnλ||u|| + ||u|| ].
p p
dφ (λu)
Let = 0 , which implies
dλ
dφ (λu) dφ (λu)
> 0 on (0 , λ2 ); < 0 on (λ2 , ∞).
dλ dλ
dJ (λu)
I ( λu ) = λ . (21.20)
dλ
Lemma 21.6. Assume that u ∈W01,p (Ω )\{0}, and ||u||pp ≠ 0. Then, we obtain
sup λu = d where
λ >0
n 2(n + p( l +1))− p 2 ( l +1)
d = [2π (l + 1)] e
2 2 .
Proof. If I(λ*u) = 0 and ||u||pp ≠ 0, from Lemma 21.4 and Equation (21.10),
we obtain
Viscoelastic Equation of p-Laplacian Hyperbolic Type 313
t
I (λ *u) ( p − 2)
sup λu = J (λ *u) =
λ >0 p
+
2 p
1 − ∫0
g (s )ds ||∇λ *u)||2
1 p 1 1
+ 2 ||λ *u )||p + − ( g °∇ (λ *u))(t )
p 2 p
( p − 2) 1
≥ l ||∇λ *u)||2 + 2 ||λ *u)||pp
2p p
1 p
≥( 2 ||λ *u )||p . (21.21)
p
2
Since 0 < < 1, now by using the following inequality
p
xv ≤ x + 1
we have
2
t
I (u) ||∇ u)|| + 1 −
=
p
p
0
∫
Ω
∫
g (s)ds ||∇ u||2 + ( g °∇ u)(t ) − (u) p ln|u| dx
( p − 2)a
2
a2
≥ ||∇ u)||pp + l ||∇ u||2 + ( g °∇ u)(t ) − ||u )| | p
p − ||(∇ u)||pp
4π 2π
n a2
+ (1 + ln a)||u||pp − ||u||pp ln ||u||p ≥ (l + 1 − (||∇ u||2 −1)
p 2π
n ( p − 2)a2 p
+ (1 + ln a) − − ln ||u||p ||u||p (21.23)
p 4π
314 Mathematics and Computer Science Volume 2
Taking a = 2π (l + 1) , we find
n 1 (l + 1)( p − 2)
I (u) ≥ 1 + ln2π (l + 1) −
p
− ln ||u||p ||u||p . (21.24)
p 2 2
Therefore, it is clear that the former case is not possible. Suppose that
I(u(t*)) = 0, then u (t*) ∈ N. By the definition of d, J(u (t*)) ≥ d was
obtained, which is in contradiction with J(u (t*)) < E(0) < d. So, we showed
that the other case is not possible as well.
Theorem 21.8. Let u0 ∈ U, u1 ∈ H 01 (Ω ). Then, Equation (21.1) with
initial boundary conditions has a global weak solution u(t ) ∈ L∞ (0 , ∞;W01,p (Ω )) with utt ∈
u(t ) ∈ L (0 , ∞;W01,p (Ω )) with utt ∈ L∞ (0 , ∞; H 01 ((Ω )) under the condition 0 < E (0) < d.
∞
Proof. Our purpose is to prove that ||∇ u||pp +l ||∇ u||2 + ||∇ u||pp is bounded
for not depending on t. By using Lemma 21.4, (21.11), definition of E(t),
and the condition 0 < E(0) < d, we obtain
1
E(t ) = ||ut ||2 + J(u(t )) ≤ E(t ) < d . (21.27)
2
u(t) ∈ U. (21.28)
t
I (u) 1 1
d > J (u) = + − 1 −
p 2 p ∫0
g (s )ds ||∇ u||2
1 p 1 1
+ 2 ||u||p + − ( g °∇ u)(t )
p 2 p
( p − 2)l 1 p
> ||∇ u||2 + ||u||p (21.29)
2p p2
2 pd p
> ||∇ u||2 and p 2d > ||u||p . (21.30)
( p − 2)l
316 Mathematics and Computer Science Volume 2
It follows from (A1), the definition of J(u), (21.11), and (21.27) that
1 1
∫ u ln|u| dx.
p
p ||ut ||2 + 2||∇ u||pp + pl ||∇ u||2 + ||u||p ≤ 2 pd + p
p p
Ω (21.31)
1 p
p ||ut ||2 + 2||∇ u||pp + pl ||∇ u||2 + ||u||p
p
( p − 2)a 2 p a2 p
≤ 2 pd + ||u||p + ||∇ u||p
4π 2π
n p p
− (1 + ln a)||u||p + ||u||p +ln ||u||p . (21.32)
p
1 1 1 1 4 pd
p ||ut ||2 + (1 − l )||∇ u||pp + pl ||∇ u||2 ≤ max , , ln p.
p 1 − l pl pl ( p − 1)
(21.34)
Viscoelastic Equation of p-Laplacian Hyperbolic Type 317
Therefore, we have solutions (u) for the Equation (21.1) that exist globally,
by using (21.34) and by the same argument, the continuation theorem [27, 28].
(21.35)
2
Since 0 < < 1, by using (21.22), we have
p
a2
1 − + l (||∇ u||2 −1) + ( g °∇ u)(t )
2π
n ( p − 2)a 2 p
+ (1 + ln a) − ln ||u||p − ||u||p
p 4π
< 0.
n ( p − 2)a 2 (21.37)
(1 + ln a) − ln ||u||p − < 0,
p 4π
G ′(t ) =
∫ 2uu dx + ||∇ u|| − ||∇ u ||
Ω
t
2
0
2
=
∫ 2uu dx + ∫ ∫ 2|∇ u(s)||∇ u (s)|dxds.
Ω
t
0 Ω
t (21.40)
Viscoelastic Equation of p-Laplacian Hyperbolic Type 319
G ′′(t ) = 2
∫ |u | dx + 2 ∫ uu dx + 2 ∫ |∇ u||∇ u |dx
Ω
t
2
Ω
tt
Ω
t
= 2||ut ||2 −2
∫ |∇ u| dx − 2 ∫ |∇ u| dx
Ω
2
Ω
p
+2
∫ ∫ g (t − s)∇ u(s)∇ u(t )dxds + 2 ∫ |u| ln|u|dx.
Ω 0 Ω
p
(21.41)
Since
(21.41) leads to
t
∫
2 p
G ′′(t ) ≥ 2||ut || −2 1 − g (s )ds ||∇ u||2 dx − 2||∇ u||p
0
t
+
∫ g (t − s)(||∇ u(s)|| )ds − 2( g °∇ u)(t ) + 2 ∫ |u| ln |u|dx
0
2
Ω
p
2
≥ 2||ut || −2 I (u). (21.42)
320 Mathematics and Computer Science Volume 2
( p 2 + 2)
G ′′G − (G ′ )2
4
( t
)
= 2G(t ) ||ut ||2 + ||∇ u||p + 1 − ∫ 0 g (s )ds ||∇ u||2 +( g °∇ u)(t ) + 2 ∫ Ω |u|p ln |u| dx
p
t
2 2
−( p + 2)(G(t ) − (T − t )||∇ u0 || ) × (||ut || + 2
∫ |∇ u (s)| ds + ( p + 2)(t ).
0
t
2 2
(21.43)
where
t t
(t ) = ||u||2 +
∫
0
|∇ u(s )| ds × ||ut ||2 +
2
∫ 0
|∇ ut (s )|2ds
t 2
−
∫ uu dx + ∫ ∫
Ω
t
0 Ω
|∇ u(s )||∇ ut (s )|dxds .
(21.44)
∫ uu dx ≤ ||u|| ||u || ,
Ω
t
2
t
2
(21.45)
t t t
0
2
0
t
2
(21.46)
and
t
2
∫ uu dx + ∫ ∫ |∇ u(s)||∇ u (s)|dxds
Ω
t
0 Ω
t
t t
≤ ||ut || 2
∫ ||∇ u(s)|| ds + ||u|| ∫ ||∇ u (s)|| ds.
0
2 2
0
t
2
(21.47)
Viscoelastic Equation of p-Laplacian Hyperbolic Type 321
p2 + 2
G ′′G − (G ′ )2 ≥ ΛG (21.48)
4
t
Λ(t ) = 2 ||ut ||2 + ||∇ u||p
∫ ∫ |u|p ln|u|dx
p
+1 − g (s )ds ||∇ u||2 + ( g °∇ u)(t ) −
0 Ω
t
− ( p + 2) ||ut ||2 +
∫
0
|∇ ut (s )|2 ds .
By using definition of total energy E(t), (A1), (A2), (21.22) and the
Logarithmic Sobolev inequality, it follows that
−2( p + 1)
∫ |u| ln|u|dx − ( p + 2)∫ |∇ u (s)| ds
Ω
p 2
0
t
2
2
a p
≥ −2 p 2 E(t ) + 2( p + 1) 1 − ||∇ u||p
2π
t
2
+( p + 2) 1 −
∫0
g (s )ds ||∇ u||2 +( p 2 + 2)( g °∇ u)(t )
( p − 2)a 2 n p
−2( p + 1) − (1 + ln a) + ln ||u||p ||u||p
4π p
t
∫ |∇ u (s)| ds + 2||u||
2 2 p
−( p + 2) t p (21.49)
0
a2 p
≥ −2 p 2 E(t ) + ( p 2 + 2) 1 − + l )(||∇ u||2 −1) + 2||u||p
2π
( p − 2)a 2 n p
−2( p + 1) − (1 + ln a) + ln ||u||p ||u||p
0
( p − 2)a 2 n p
−2( p + 1) − (1 + ln a) + ln ||u||p ||u||p
4π p
t
∫
2 2 p
322 Mathematics |∇ utComputer
−( p + 2) and (s )| ds + 2||u||p Volume 2
Science
0
a2 p
≥ −2 p 2 E(t ) + ( p 2 + 2) 1 − + l )(||∇ u||2 −1) + 2||u||p
2π
( p − 2)a 2 n p
−2( p + 1) − (1 + ln a) + ln ||u||p ||u||p
4π p
t
2
−( p + 2)
∫ |∇ u (s)| ds.
0
t
2
t t
2
≥ −2 p E(0) −
∫ 0
|∇ ut (s )| ds + 2 p 2d − ( p 2 + 2)
2
∫ |∇ u (s)| ds
0
t
2
= −2 p 2 E(0) + 2 p 2d + ( p 2 − 2)
∫ |∇ u (s)| ds.
0
t
2
(21.50)
Λ(t) ≥ 0. (21.51)
p2 + 2
G′′G − (G′)2 ≥ 0. (21.52)
4
Viscoelastic Equation of p-Laplacian Hyperbolic Type 323
p2 + 2
ln|G ′(t )| − ln|G(t )| 4 ≥ 0. (21.53)
G(0)
G(t ) ≥ 4 . (21.54)
p2 − 2 p2 − 2 2
p −2
1 − G(0) 4 t
4
p2 − 2
G ′(0) 4
From definition of G(t), it is clear to conclude that < 1 where
G(0)
G(0) =|| u ||2 + T || ∇u ||2 . Consequently, we can rewrite (21.54) so that
G(0)
G(t ) ≥ 4 .
2
( p − 2) G ′(0) p2 − 2
1 − 4 G(0)
2||u||2
T≤
( p 2 − 2) ∫ uu dx − 2||∇ u || .
Ω
0t 0
2
References
1. Adams, R. A., Fournier, J. J. Sobolev spaces. Pure and applied mathematics
(Amsterdam), (2003).
2. Al-Gharabli, M. M., Al-Mahdi, A. M., Messaoudi, S. A. Decay Results for a
Viscoelastic Prob- lem with Nonlinear Boundary Feedback and Logarithmic
Source Term. Journal of Dynamical and Control Systems, 1-19, (2020).
3. Andrade, D., Silva, M. J., Ma, T. F. Exponential stability for a plate equation
with p-Laplacian and memory terms. Mathematical Methods in the Applied
Sciences, 4(35), 417-426, (2012).
4. Araújo, G. M., Araújo, M. A. F., Pereira, D. C.On a variational inequality for
a plate equation with p-Laplacian end memory terms. Applicable Analysis,
1-14, (2020).
5. Bialynicki-Birula, I., Mycielski, J. Wave equations with logarithmic nonlin-
earities. Bull. Acad. Pol. Sci. Cl, 3(23), 461-466, (1975).
6. Bouhali, K., Ellaggoune, F. Viscoelastic wave equation with logarithmic non-
linearities in Rn. J Partial Differ Equ, 30(1), 47-63, (2017).
7. Buljan, H., Šiber, A., Soljačíc, M., Schwartz, T., Segev, M., Christodoulides,
D. N. Incoherent white light solitons in logarithmically saturable noninstan-
taneous nonlinear media. Physical Review E, 68(3), 036607, (2003).
8. Cazenave, T., Haraux, A. Équations d’évolution avec non linéarité logarith-
mique. In Annales de la Faculté des sciences de Toulouse: Mathématiques,
2(1), 21-51, (1980).
9. Di, H., Shang, Y., Song, Z. Initial boundary value problem for a class of
strongly damped semilinear wave equations with logarithmic nonlinearity.
Nonlinear Analysis: Real World Ap- plications, 51, 102968, (2020).
10. Ding, H., Wang, R., Zhou, J. Infinite time blow-up of solutions to a class of
wave equations with weak and strong damping terms and logarithmic non-
linearity. Studies in Applied Mathematics, 147(3), 914-934, (2021).
11. Ferreira, J., Irkıl, N., Raposo, C. A. Blow up results for a viscoelastic
Kirchhoff-type equation with logarithmic nonlinearity and strong damping.
Mathematica Moravica, 25(2), 125-141, (2021).
12. Ha, T. G., Park, S. H. Blow-up phenomena for a viscoelastic wave equation
with strong damping and logarithmic nonlinearity. Advances in Difference
Equations, 2020 (1), 1-17, (2020).
13. Liu, W., Yu, J. On decay and blow-up of the solution for a viscoelastic wave
equation with boundary damping and source terms. Nonlinear Analysis:
Theory, Methods Applications, 74(6), 2175-2190, (2011).
14. Liu, W., Li, G., Hong, L. Decay of solutions for a plate equation with
p-Laplacian and memory term. Electronic Journal of Differential Equations,
2012(129), 1-5, (2012).
15. Liu, C., Ma, Y. Blow up for a fourth order hyperbolic equation with the loga-
rithmic nonlinearity. Applied Mathematics Letters, 98, 1-6, (2019).
Viscoelastic Equation of p-Laplacian Hyperbolic Type 325
16. Ma, L., Fang, Z. B. Energy decay estimates and infinite blow-up phenomena
for a strongly damped semilinear wave equation with logarithmic nonlinear
source. Mathematical Methods in the Applied Sciences, 41(7), 2639-2653,
(2018).
17. De Martino, S., Falanga, M., Godano, C., Lauro, G. Logarithmic Schrödinger-
like equation as a model for magma transport. EPL (Europhysics Letters),
63(3), 472, (2003).
18. Park, S. H. Stability for a viscoelastic plate equation with p-Laplacian. Bulletin
of the Korean Mathematical Society, 52 (3), 907-914, (2015).
19. Pereira, D. C., de Araújo, G. M., Raposo, C. A.. Unilateral Problem for a
Viscoelastic Beam Equation Type p-Laplacian with Strong Damping and
Logarithmic Source. European Journal of Mathematical Analysis, 2 (5), 1-20,
(2022).
20. Peyravi, A. General stability and exponential growth for a class of semi-
linear wave equations with logarithmic source and memory terms. Applied
Mathematics and Optimization, 81(2), 545-561, (2020).
21. Del Pino, M., Dolbeault, J., Gentil, I. Nonlinear diffusions, hypercontractiv-
ity and the optimal Lp-Euclidean logarithmic Sobolev inequality. Journal of
Mathematical Analysis and Applications, 293(2), 375-388, (2004).
22. Pişkin, E., Okutmuştur, B. An Introduction to Sobolev Spaces, Bentham
Science, (2021).
23. Pişkin, E., Irkıl, N. Well-posedness results for a sixth-order logarithmic
Boussinesq equation. Filomat, 33(13), 3985-4000, (2019).
24. Pişkin, E., Boulaaras, S., Irkil, N. Qualitative analysis of solutions for the p-
Laplacian hyperbolic equation with logarithmic nonlinearity. Mathematical
Methods in the Applied Sciences, 44(6), 4654-4672, (2021).
25. Irkıl, N., Pişkin, E. Global Existence and Decay of Solutions for a Higher-
Order Kirchhoff- Type Systems with Logarithmic Nonlinearities. Quaestiones
Mathematicae, 1-24, (2021), (in press).
26. Raposo, C. A., Cattai, A. P., Riberio, J. O.Global solution and asymptotic
behaviour for a wave equation type p-Laplacian with memory. Open Journal
of Mathematical Analysis, 2, 156-171, (2018).
27. Segal, I. Nonlinear semigroups. Annals of Mathematics, 339-364, (1963).
28. Georgiev, V., Todorova, G. Existence of a solution of the wave equation with
nonlinear damping and source terms. Journal of differential equations, 109
(2), 295-308, (1994).
29. M. Willem, Minimax Theorems. Progress in Nonlinear Differential Equations
and Their Applications 24, Birkhouser Boston Inc. Boston, MA, (1996).
30. Xu, R., Lian, W., Kong, X., Yang, Y. Fourth order wave equation with non-
linear strain and logarithmic nonlinearity. Applied Numerical Mathematics,
141, 185-205, (2019).
31. Ye, Y. Logarithmic viscoelastic wave equation in three-dimensional space.
Applicable Analysis, 100(10), 2210-2226, (2021).
326 Mathematics and Computer Science Volume 2
2
Tribhuvan University, Central Department of Mathematics, Kathmandu, Nepal
Abstract
The dynamics of network flows are influenced by transit time along the arcs, unlike
the static cases. A continuous-time model depicts the actual scenario of many
transshipment problems. Here, we introduce the average arc problem (AAP) in a
continuous-time model which widely occurs in our daily life. In AAP, the flow value
which is total of the flow at every time step/instant (in discrete/continuous time
respectively) of the arc has an upper bound as the flow capacity of that arc. When
the time-expanded network is structured, AAP violates the capacity constraint. This
failure of extension of the arc problem (AP) to the AAP of the problem leads to for-
mulating a linear programming of it. In AP, at every instant of time, flow can be sent
up to the capacity of that arc, as in the traditional dynamic network flows. To get
feasible flow, a transformation and reduction algorithm is also presented.
22.1 Introduction
Consider the water supply network of a city, the traffic network of a high-
way connecting two cities, or the communication network between two
places. In these type of network flows, the capacity of the arc in every time
step in discrete time (in every instant in continuous-time) can be fulfilled
by the corresponding flows. Ford and Fulkerson [6] introduced such flow
dynamics by making dynamic networks with a source and destination.
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (327–336) © 2023 Scrivener Publishing LLC
327
328 Mathematics and Computer Science Volume 2
The algorithm with polynomial running time was firstly studied by Hoppe
and Tardos [10] and solved flow sending problems in dynamic cases hav-
ing more than one source and/or sink.
In such cases, if the transit time and capacity of an arc are 5 and 3 units
(say) respectively, then each time step can contain 3 units of flow, as shown
in Figure 22.1. Here, the capacity is not fulfilled in time steps 1 and 4. We
refer to this type of flow problem as the arc problem (AP).
In our paper, we have switched to the cases where some restrictions in
each time steps flow cannot be fulfilled. Total value of the flow of all time
steps of the transit time of an arc is bounded above by the capacity of that
arc (see Figure 22.2), where total of the flow value of time steps 1, 3, and 4
is 3, the capacity of that arc. We call this problem the average arc problem
(AAP).
The structure of this paper is as follows. Section 22.2 gives background
on how flow dynamics work in a discrete as well as continuous-time set-
ting along with the short view on AAP. In Section 22.3, failure of extension
of AP to AAP is discussed with the help of a time-expanded network. A
mathematical formulation for the AAP and transformation and reduction
algorithm are introduced in Section 22.4. Lastly, the paper is wrapped with
conclusion in Section 22.5.
0 1 2 3 4 5
0 1 2 3 4 5
Figure 22.2 Average arc with transit time 5 and capacity 3 units.
Flow Dynamics in Continuous-Time with Average Arc Capacities 329
capacity, per time unit storage cost per unit flow, and supply/demand rate
at vertex or node i, at time t.
Let Fi(t) be the flow rate from supersource ssi to source si(from sink di to
supersink sdi) at time t. Also, let Fi be the flow amount supplied from super-
supersource sss to supersource ssi (from supersink sdi to supersupersink ssd).
sss
ss1 ss2
m=6 s1 s2 m = 4
2
1 1
1
x y
2 1
d m = 10
sd
ssd
Figure 22.3 Simple instance of dynamic network of an arc problem, arc with transit time,
and capacity 1 unit each.
s1 s2 x y d 3
ss2 s1 s2 x y d [2, 3)
sss sd ssd
ss1 s1 s2 x y d [1, 2)
s1 s2 x y d [0, 1)
Consider the flow through the arc (s1, y) in time [0, 1) and [1, 2] and
the total flow value is 1 + 1 = 2 units, which violates the capacity of flow in
AAP which leads to infeasibility of flow and hence the failure in extension
of AP to AAP.
22.4 Formulation
Using the notations introduced above, the minimum time of dynamic flow
problem (MTP) suggested by Philpott [16] is:
MTP:
min T (22.1)
∫ ∑ ∫ ∑ f (t −τ ) + s (t ) =
t t
s.t . fij (t )dt − ji s (0) ji i i (22.2)
0 0
∫ ∑ ∑ (t + τ
T
min id ) fid (t ) dt (22.4)
0
i →d t∈[0,T ]
t +τ ij
∫ ∑ f (t ) dt ≤ u
t
ij ij (22.5)
0
t
Fi(t) ≤ mi (22.7)
Fi ≤ mi (22.8)
∫ ∑ ∫ ∑f
t t
fij (t − τ ij ) dt = jk (t ) dt (22.9)
0 0
i→ j j→k
332 Mathematics and Computer Science Volume 2
∫ ∑ ∫ ∑f
t t
− τ ij ) dt + s j (0)
fij (t= jk (t ) dt + s j (t ) (22.10)
0 0
i→ j j→k
∫ ∑ f (t ) dt
t
Fs (t ) = sj (22.11)
0
s→ j
∫ ∑ f (t −τ
t
=Fd (t ) id id ) dt (22.12)
0
i →d
∫ ∑ F (t ) dt
T
Fi = i (22.13)
0
i∈ssi /sdi
∑F
s∈sources
s = ∑m
s∈sources
s (22.14)
∑F
d∈sinks
d = ∑m
d∈sinks
d (22.15)
The linear objective function 22.4 aims to minimize the arrival time
of the flows at the sink d. Similarly, 22.5, 22.6, 22.7, and 22.8 give capac-
ity constraints of the flows, 22.9 is the flow conservation constraint for
intermediate vertex j, 22.10 expresses flow non-conservation constraint,
22.11 and 22.12, respectively, are the conservation constraints for flow in
the sources and sinks, 22.13 is the flow conservation constraints for the
supersources and supersinks, and lastly 22.14 and 22.15 are the total mass
supplied/demanded from supersupersource/supersupersink.
The feasibility of AAP and the above given linear formulation are
bi-conditional to each other. To observe the feasibility of flow, the concept
of PARTITION is used.
PARTITION Problem
Consider a finite set which has n items with sizes b1, ..., bn ∈ Z+, such that
∑nk =1 bk =
2 J , for some J ∈ Z+.
Does some subset I ⊂ {1, ..., n} satisfy the property ∑ k∈I bk = J?
Flow Dynamics in Continuous-Time with Average Arc Capacities 333
Theorem 22.1 [13] The necessary and sufficient condition to send two units
of dynamic flow from s to d in time T is the existence of a ‘yes’-instance
PARTITION problem.
The idea to prove the sufficient part of Theorem 22.1 is the flow of two
units is sent along two paths having a disjointed arc in two packets, each
having one unit of flow and length n and this can be done in both cases
within the prescribed time horizon T. For the necessary part, firstly
an arbitrary arc is taken which has more than one unit flow value and
total time required to reach the flow to the sink d is calculated which
yields more than the time horizon T, which leads to the contradiction.
Again, if a supposition is made against the yes instance PARTITION, i.e.,
∑ k∈I bk = J + ∈, for some ϵ ≠ 0, then total transit time of the flow on one
of the two paths will exceed T.
If an NP-complete problem fails to be NP-complete after satisfying sim-
ilarity assumption, then it is called weakly NP-complete. Using the concept
of PARTITION and the above theorem, Melkonian [13] showed AAP is
NP-complete in the weak sense. To reduce AAP to AP, only the capacities
of the arcs are modified, dividing by the corresponding transit times of
the arcs where capacity constraints are violated. The following algorithm
performs well for the feasibility of flow with the time optimization given
by the AAP formulation and its execution steps ahead on the basis of the
length of arc having the least transit time.
- Input: Given an AP with N = (V, A, T, fij, uij, wij, τij, si, vi, xi, mi)
22.5 Conclusion
When AP is transformed into the time-expanded network, it violates
capacity constraints in AAP. This infeasibility of flow leads to the linear
formulation of AAP. Using the well-known NP-complete PARTITION
problem, AAP is reduced to AP by some modification on arc capacities, i.e,
averaging the capacities by the corresponding transit times. The feasibility
of reduced AP can be obtained by the algorithm introduced by Hoppe and
Tardos [10], implying the feasibility of the original AAP. Practically, the
hybrid network would be functioning simultaneously. After observing the
feasibility of the flow, the optimality can be obtained by the suitable opti-
mality seeking method (such as network simplex method, etc.).
Acknowledgment
The first author acknowledges University Grants Commission, Nepal for
PhD fellowship.
References
1. E. J. Anderson, P. Nash and A. B. Philpott (1982). A class of continuous net-
work flow problems. Mathematics of Operation Research, 7: 501-514.
2. J. E. Aronson (1989). A survey of dynamic network flows. Annals of
Operations Research, 20: 1-66.
3. N. Baumann and M. Skutella (2006). Solving evacuation problems effi-
ciently–earliest arrival flows with multiple sources. In: 47th annual IEEE
symposium on foundations of computer science (FOCS’06), 399–410.
4. R. Burkard, K. Dlaska and B. Klinz (1993). The quickest flow problem. ZOR
Meth Models Oper Res, 37: 31–58.
5. L. R. Ford and D. R. Fulkerson (1958). Constructing maximal dynamic flows
from static flows. Operations Research Letters, 6: 419-433.
6. L. R. Ford and D. R. Fulkerson (1962). Flows in networks. Princeton
University Press, New Jersey.
7. D. Gale(1959). Transient flows in networks. Michigan Math J, 6(1): 59–63.
8. H. Hamacher and S. Tjandra (2001). Mathematical modeling of evacuation
problems: a state of art. Berichte des Frauenhofer ITWM, Nr. 24. http://
www.itwm.fraunhofer.de/fileadmin/ITWMMedia/Zentral/Pdf/Berichte
ITWM/2001/bericht24.pdf. Accessed 18 Oct 2006.
9. B. Hoppe and E. Tardos (1994). Polynomial time algorithms for some evacu-
ation problems. In: Proceedings of the fifth annual ACM-SIAM symposium
Flow Dynamics in Continuous-Time with Average Arc Capacities 335
Bihar, India
2
Department of Mechanical Engineering, BIT Sindri, Jharkhand, India
Abstract
In the present work, a M/M/3 Queuing system with a Queue-dependent multi-server
has been considered. Here, a number of failed machines form a queue and repair-
men consider a service provider or service channel which starts its service when
the queue length is N. We found a generating function for breakdown machines.
Afterwards, some performance measures including idle time and busy time for the
system have been evaluated. At last, cost is optimized using a genetic algorithm.
Keywords: M/M/3, repairmen, queue dependent server, genetic algorithm
23.1 Introduction
The concept of a multiserver queueing system with queue-dependent serv-
ers is not new. Hahn and Sivazlian, in 1990, developed the M/M/2 model
with service stations [5]. Yadin and Naor, in 1963, invented the N policy
concept for the single server system [9]. The (0, K, N, M) rule in a two
server system used by Rhee and Sivazlian in 1990 helps to derive the work-
ing period distribution [12].
Natarajan studied a system for warm stand-bys in 1968 [13]. With one
online unit and two standbys with failure and repair time exponentially
distributed, this model assumes only one condition, namely that the sys-
tem will fail if spares are not present for the breakdown machine. There is
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (337–348) © 2023 Scrivener Publishing LLC
337
338 Mathematics and Computer Science Volume 2
an analytical closed-form solution for the M/Ek/1 system [15]. The server
system for standbys and 2 types of failure time distribution is also examined.
Wang and Ke (2000) introduce a supplementary and recursive technique
for developing a solutions to a single server queueing system working with
the N policy [16]. The current system is that the line length determines the
server value in the system. In the event that no machines fail in the system,
all the fixers are temporarily inactive until the length of the line reaches a
predetermined level, i.e., W, which is also a decision change. For this sin-
gle server period, it will be activated instantly. Sometime later, if the line
length in the system goes up to the aforementioned level, it says S, which is
less than W and then the other available server will also run immediately.
Similarly, if the failure machines in the system go up to a next level, say T,
which is less than S, then all the servers will be activated immediately.
The objectives of this paper are:
23.3 Notations
ϸ[0, n] = Occurrence of n customers in the model where there are
no operators, when n ϵ [0, W-1]
ϸ[1, n] = Occurrence of n customers in the model when one mod-
ifier works, when n ϵ [0, S-1]
ϸ[2, n] = Occurrence of n customers in the model, where two
moderators are working, when n ϵ [q + 1, T-1]
Analysis of a Multiserver System of Queue-Dependent Channel 339
[1 − n ]
[1,n] = [0,0] , n [1, q] (23.23)
(1 − )
S−W −1 W
n−Q
(1 − ) 1 −
[ 2 ,n ] = 2 , n [q +1, S] (23.26)
[0,0]
(2 − )(1 − S−Q )
342 Mathematics and Computer Science Volume 2
n−S S−Q
S−W −1 W
(1 − ) 1 −
2 2
[ 2 ,n ] = [0 ,0] , n [S, S + 1]
(2 − )(1 − S−Q )
(23.27)
T−S−1 S
n−r
(1 − ) 1 −
[3 ,n] = 3 , n [r +1,T ] (23.28)
[0 ,0]
(3 − )(1 − ρ T−r )
n −T T−r
T−S−1 S
(1 − ) 1 − [0 ,0]
3 3
[ 3 ,n ] = , n [ T + 1, L] (23.29)
(3 − )(1 − T−r )
α
Where =
β
T −1
W −1
S −1
T −1
The queue length of a busy one-time server one, busy two-time server, and
busy three-time server are calculated. Since a working period is the addi-
tion of the leisure time, the busy time of 1 fixer, the busy time of 2 fixers,
and the busy time of 3 fixers, EC = EB1 + EB2 + EB3 + EI
EI EB
= min{ Ch Ls + (CV 1 + CV 2 + CV 3 ) + C01 1
EC EC
EB2 EB
+ C02 + C03 3 + α Cl P(3,l )}
EC EC
Case I: Ch = 100, C01 = 100, C02 = 200, C03 = 250, Cv = Cv1+ Cv2+
Cv3 = 300
Case II: Ch = 150, C01 = 100, C02 = 200, C03 = 300, Cv = Cv1+ Cv2+
Cv3 = 300
Case III: Ch = 100, C01 = 150, C02 = 250, C03 = 300, Cv = Cv1+ Cv2+
Cv3 = 500
Case IV: Ch = 100, C01 = 100, C02 = 200, C03 = 300, Cv = Cv1+ Cv2+
Cv3 = 400
Case V: Ch = 100, C01 = 100, C02 = 200, C03 = 300, Cv = Cv1+ Cv2+
Cv3 = 300
Case VI: Ch = 100, C01 = 100, C02 = 250, C03 = 300, Cv = Cv1+ Cv2+
Cv3 = 400
Case VII: Ch = 100, C01 = 100, C02 = 200, C03 = 400, Cv = Cv1+
Cv2+ Cv3 = 200
For all the costs listed above, optimal value and expected cost for dif-
ferent values of α are shown in Table 23.1. It is clearly seen that total cost
increases as α increases. When all the servers are on vacation, Cv increases
(See Case III & Case VII) as it includes the cost associated with Cl and
346 Mathematics and Computer Science Volume 2
Table 23.1 Expected cost E (q*, r*, W*, S*, T*) and optimal value (q*, r*, W*, S*, T*) (For α=4, 6, 7, 8, 9, 10).
α 4 6 7 8 9 10
I q*, r*, W*, S*, T* 1,3,4,5,6 2,3,4,5,6 1,2,3,4,6 2,3,4,6,7 4,5,6,7,8 3,4,5,6,7
E(q*, r*, W*, S*, T*) 1099.89 1001.43 1003.90 1005.87 1006.34 1007.21
II q*, r*, W*, S*, T* 2,3,4,5,6 1,2,3,4,6 3,4,5,6,7 2,3,4,5,8 3,4,5,6,8 4,5,6,7,8
E(q*, r*, W*, S*, T*) 1424.34 1461.45 1495.74 1501.21 1503.90 1505.43
III q*, r*, W*, S*, T* 2,3,4,6,7 3,4,5,6,8 1,2,4,6,7 2,4,6,7,8 4,5,7,8,9 5,6,7,8,9
E(q*, r*, W*, S*, T*) 2007.39 2018.14 2100.56 3112.71 3413.20 3740.38
IV q*, r*, W*, S*, T* 3,5,6,7,8 2,4,6,7,8 2,3,4,6,7 4,5,6,7,9 5,6,7,8,9 6,7,8,9,10
E(q*, r*, W*, S*, T*) 1004.13 1007.67 1011.31 1149.74 1179.89 1200.06
V q*, r*, W*, S*, T* 3,4,7,8,9 4,5,6,7,9 2,5,7,8,9 6,7,8,9,10 4,6,8,9,10 5,7,8,9,10
E(q*, r*, W*, S*, T*) 1262.11 1342.03 1376.40 1428.29 1501.10 1503.29
VI q*, r*, W*, S*, T* 4,5,6,7,8 3,4,5,7,9 4,5,7,8,10 5,6,7,9,10 6,7,8,9,10 5,7,8,9,10
E(q*, r*, W*, S*, T*) 1004.52 1007.21 1191.32 1134.69 1187.74 1200.54
VII q*, r*, W*, S*, T* 5,7,8,9,10 6,7,8,9,10 4,6,8,910 5,7,8,9,10 4,5,8,9,10 5,6,7,8,10
E(q*, r*, W*, S*, T*) 1236.00 1332.39 1737.20 2429.31 2519.93 2739.71
Analysis of a Multiserver System of Queue-Dependent Channel 347
holding cost of all failed machines in the system. Cost associated with
higher threshold value is always increased.
23.5 Conclusions
The paper concluded the steady-state results for a multiserver queueing
system with fixed capacity. A cost analysis is done to find the best level (q*,
r*, W*, S*, T*) of breakdown machines in the system using a genetic algo-
rithm. The flexibility of inclusion of servers increases the system efficien-
cies and avoids customer loss due to service delay. Cost analysis at different
system parameters is useful for researchers and engineers.
References
1. Anupama and Solanki, A., “Performance Analysis And Optimization Of
An M/M/3 Queuing System Under Tetrad Policy” International Journal of
Advanced Technology in Engineering and Science (IJATES), ISSN: 2348 –
7550, ISSN (online): 2348 – 7550, Vol.02(11), pp 433-444, November 2014.
2. Bell CE (1971),” Characterization and computation of optimal policies for
operating an M/G/1 queueing system with removable server”. Operations
Research 19:2008–218.
3. Bell CE (1972).,” Optimal operation of an M/G/1 priority queue with remov-
able server”. Operations Research 21:1281–1289.
4. Gnedenko et al. (1969),” Mathematical methods of Reliability Theory”
Academic press, NewYork.
5. Hahn and Sivazlian (1990),” Distribution of the busy period in a control-
lable M\M\2 under the Triadic (0‚K‚N‚M) policy”, J. Appl. Prob., Vol. 27,
pp.425-432.
6. K. H. Wang and B. D. Sivazlian (1989),” Reliability of a system with warm
standbys and repairmen”. Microelectron. Reliab, 29, 849-860.
7. Kuo-Hsiung Wang and Ya-Ling Wang (2002),” Optimal control of an M/M/2
queueing system with finite capacity operating under the triadic (0‚Q‚N‚M) pol-
icy” Math Meth Oper Res 55:447–460.
8. M. N. Gopalan (1975),” Probablistic analysis of single server n-unit system
with (n-1) warm standbys.” Oper Res. 23, 591-595.
9. M. Yadin and P. Naor (1963),” Queueing systems with a removable service
station”. Opl Res. Q. 14, 393-4005.
10. M. J. Magazine (1971),” Optimal control of multi-channel service systems”.
Naval Res. Logist. Q. 177-183.
11. M. J. Sobel (1969),” Optimal average-cost policy for a queue with start-up and
shut-down costs”. Opns Res. 17, 145-162.
348 Mathematics and Computer Science Volume 2
12. Rhee H-K, Sivazlian BD (1990),” Distribution of the busy period in a control-
lableM/M/2 queue operating under the triadic (0,K,N,M) policy”. Journal of
Applied Probability 27:425–432.
13. R. Natarajan (1968),” A reliability problem with spares and multiple repair
facilities”. Opns Res. 16, 1041-1057.
14. Subramanian et al. (1976),” Reliability of a repairable system with standbys
failure.” Opns res. 24, 169-176.
15. Wang K-H, Huang H-M (1995),”Optimal control of a removable server in an
M/Ek/1 queueing system with finite capacity”. Microelectronics and Reliability
35:1023–1030.
16. Wang K-H, Ke J-C (2000),” A recursive method to the optimal control of an
M/G/1 queueing system with finite capacity and infinite capacity”. Applied
Mathematical Modelling 24:899–914.
24
An Approach to Ranking of Single
Valued Neutrosophic Fuzzy Numbers
Based on (α, β, γ) Cut Sets
Sunayana Saikia *
Abstract
In most real-world applications, single-valued neutrosophic fuzzy numbers are
employed to make judgments in a fuzzy environment. As a result, single-valued
neutrosophic fuzzy number ranking is crucial. This paper proposes a novel
approach for ranking single-valued neutrosophic fuzzy numbers. The approach
was created by combining the α, β, and γ cuts of neutrosophic fuzzy numbers in a
convex way. In addition, the proposed approach has been used to prove a number
of theorems. This paper also offers a number of numerical examples.
Keywords: Single valued neutrosophic fuzzy numbers, α-cut set, β-cut set,
γ-cut set, ranking of neutrosophic fuzzy numbers, convex combination
24.1 Introduction
Real-life phenomena cannot be classified by real numbers due to the exis-
tence of linguistic concepts. Fuzzy numbers can be used to rate such issues
because fuzzy numbers may deal with verbal words that are ambiguous or
vague. As a result, numerous ways for ranking fuzzy numbers have been
presented at various times. In addition, the ordering of fuzzy numbers has
been thoroughly examined. In the literature, there are several research arti-
cles that describe various strategies for sorting fuzzy numbers. In a nutshell,
no universally accepted method for sorting fuzzy numbers exists. Existing
E-mail: sunayana.saikia@gmail.com
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (349–362) © 2023 Scrivener Publishing LLC
349
350 Mathematics and Computer Science Volume 2
0, if x < a ,
x − a , if a ≤ x ≤ b ,
b − a
µ A (x ) = w A , if b ≤ x ≤ c ,
d − x
, if c ≤ x ≤ d ,
d − c
0, if x > d.
w A ( x − a)
b−a , if a ≤ x ≤ b ,
wA , if b ≤ x ≤ c ,
µ(x ) =
w A (d − x ) , if c ≤ x ≤ d ,
d −c
0, otherwise.
(q − x ) + u A (x − p)
, if p ≤ x ≤ q ,
q− p
uA , if q ≤ x ≤ r ,
v( x ) =
( x − r ) + u A (s − x ) , if r ≤ x ≤ s ,
s −r
1, otherwise.
(i) A is neutrosophic-normal,
(ii) A is neutrosophic-convex,
(iii) µA is upper semi-continuous, νA, and ρA are lower semi-
continuous,
(iv) Support (S(A)) of A is bounded, i.e., S(A) = {µA(x) > 0, νA < 1,
ρA < 1, ∀x ∈ X}”.
Definition 24.2.10 ([17]). “A SVNS, A = {⟨x, µA(x), νA(x), ρA(x)⟩∀x ∈
X} = ⟨(a1, b1, c1), (i1, j1, k1), (p1, q1, r1)⟩ in the set of real numbers is said
to be single-valued triangular neutrosophic number (SVTNN), such that
p1 ≤ i1 ≤ a1 ≤ q1 ≤ j1 ≤ b1 ≤ c1 ≤ j1 ≤ r1 are the parameters. Then, the truth
membership degree, the indeterminacy membership degree, and thee fal-
sity membership degree of A can be defined as:”
x − a1
b − a , if a1 ≤ x ≤ b1 ,
1 1
1, if x = b1 ,
TA (x ) =
c1 − x , iff b1 < x ≤ c1 ,
c1 − b1
0, otherwise.
j1 − x
j − i , if i1 ≤ x ≤ j1 ,
1 1
0 , if x = j1 ,
I A (x ) =
x − j1 , iff j1 ≤ x ≤ k1 ,
k1 − j1
1, otherwise.
354 Mathematics and Computer Science Volume 2
p1 − x
q − p , if p1 ≤ x ≤ q1 ,
1 1
0, if x = q1 ,
FA (x ) =
x − j1 , iff j1 ≤ x ≤ k1 ,
k1 − j1
1, otherwise.
1 1
=
∫
0
α
∫
λ1 ,λ2 ,λ3 ( A) λ1 f (α )L ( A)dα + (1 − λ1 ) f (α )Rα ( A)dα
0
1 1
∫ ∫
+ λ2 g (β )L ( A)d β + (1 − λ2 ) g (β )R β ( A)d β
β
0 0
1 1
∫ ∫
+ λ3 h(γ )Lγ ( A)dγ + (1 − λ3 ) h(γ )Rγ ( A)dγ
0 0
24.4 Theorems
Some crucial and fundamental theorems are discussed here.
Theorem 24.4.1. Let λ1 ,λ2 ,λ3 be linear.
Proof. Let, A = ⟨(a1, b1, c1), (i1, j1, k1), (p1, q1, r1)⟩ and B = ⟨(a2, b2, c2), (i2, j2,
k2), (p2, q2, r2)⟩ be two SVTNNs and κ be any real number. Then, it has to
prove that λ1 ,λ2 ,λ3 ( A + κ B) =
λ1 ,λ2 ,λ3 ( A) + κ λ1 ,λ2 ,λ3 (B). Then,
1 1
λ1 ,λ2 ,λ3 ( A + κ B) is equal to λ1 ∫ f (α )Lα ( A + κ B)dα + (1 − λ1 ) ∫ f (α )
0 0
1 1
Rα ( A + κ B)dα + λ2 ∫ g (β )Lβ ( A + κ B)d β + (1 − λ2 ) ∫ g (β )R β ( A + κ B)d β + λ3
0 0
1 1 1
∫ h(γ )Lγ ( A + κ B)dγ + (1 − λ3 ) ∫ h(γ )Rγ ( A + κ B)dγ ,which leads to λ1 ∫
0 0 0
1 1
f (α )(Lα ( A) + κ Lα (B))dα + (1 − λ1 ) ∫ f (α )(Rα ( A) + κ Rα (B))dα + λ2 ∫ g (β )
0 0
1 1
(Lβ ( A) + κ Lβ (B))d β + (1 − λ2 ) ∫ g (β )(R β ( A) + κ R β (B))d β + λ3 ∫ h(γ )(Lγ ( A) +
0 0
1 1
κ Lγ (B))dγ + (1 − λ3 ) ∫ h(γ )(Rγ ( A) + κ Rγ (B))dγ , then it is true that(λ1 ∫ f (α )
0 0
1 1 1
Lα ( A)dα + (1 − λ1 ) ∫ f (α )Rα ( A)dα + λ2 ∫ g (β )Lβ ( A)d β + (1 − λ2 ) ∫ g (β )R β
0 0 0
1 1 1
( A)d β + λ3 ∫ h(γ )Lγ ( A)dγ + (1 − λ3 ) ∫ h(γ )Rγ ( A)dγ ) + κ (λ1 ∫ f (α )Lα (B)dα +
0 0 0
3 1
0 0 0
1 1
f (α )(Lα ( A) + κ Lα (B))dα + (1 − λ1 ) ∫ f (α )(Rα ( A) + κ Rα (B))dα + λ2 ∫ g (β )
0 0
1 1
(Lβ ( A) + κ Lβ (B))d β + (1 − λ2 ) ∫ g (β )(R β ( A) + κ R β (B))d β + λ3 ∫ h(γ )(Lγ ( A) +
0 0
Ranking of Single
1
Valued Neutrosophic Fuzzy Numbers 1 357
κ Lγ (B))dγ + (1 − λ3 ) ∫ h(γ )(Rγ ( A) + κ Rγ (B))dγ , then it is true that(λ1 ∫ f (α )
0 0
1 1 1
Lα ( A)dα + (1 − λ1 ) ∫ f (α )Rα ( A)dα + λ2 ∫ g (β )Lβ ( A)d β + (1 − λ2 ) ∫ g (β )R β
0 0 0
1 1 1
( A)d β + λ3 ∫ h(γ )Lγ ( A)dγ + (1 − λ3 ) ∫ h(γ )Rγ ( A)dγ ) + κ (λ1 ∫ f (α )Lα (B)dα +
0 0 0
1 1 1
(1 − λ1 ) ∫ f (α )Rα (B)dα + λ2 ∫ g (β )Lβ (B)d β + (1 − λ2 ) ∫ g (β )R β (B)d β
0 0 0
1 1
+λ3 ∫ h(γ )Lγ (B)dγ + (1 − λ3 ) ∫ h(γ )Rγ (B)dγ ), then the result follows
0 0
that λ1 ,λ2 ,λ3 ( A) + k λ1 ,λ2 ,λ3 (B).
24.6 Conclusion
The majority of real-life decision-making is based on fuzzy numbers due
to their language connotations. Ordering fuzzy numbers is much more
important in order to solve such difficulties. There are several documented
ranking systems for single-valued fuzzy numbers. This paper developed a
novel ranking approach for single-valued fuzzy numbers based on convex
combinations of the fuzzy numbers’ α, β, and γ cut sets. A number of the-
orems have also been proved using the proposed method. Furthermore,
numerical examples have been provided to demonstrate the suggested
strategy’s superior performance. This ranking method may be used to look
at the rankings of intuitionistic fuzzy numbers, type-2 fuzzy numbers,
hesitant fuzzy numbers, and other fuzzy numbers. Further, this proposed
Ranking of Single Valued Neutrosophic Fuzzy Numbers 359
References
1. S. I. A. Aal, M. M. A. A. Ellatif, and M. M. Hassan. Two Ranking Methods
of Single Valued Triangular Neutrosophic Numbers to Rank and Evaluate
Information Systems Quality, 2018.
2. A. Bhaumik, S. Roy, and D.-F. Li. (α, β, γ)-cut set based ranking approach
to solving bi-matrix games in neutrosophic environment. Soft Computing,
25:1–11, 02 2021.
3. P. Biswas, S. Pramanik, and B. Giri. Value and ambiguity index based rank-
ing method of single-valued trapezoidal neutrosophic numbers and its appli-
cation to multi-attribute decision making. Neutrosophic Sets and Systems,
12:127–138, 07 2016.
4. G. Bortolan and R. Degani. A review of some methods for ranking fuzzy
subsets. Fuzzy Sets and Systems, 15(1):1 – 19, 1985.
5. M. Brunelli and J. Mezei. How different are ranking methods for fuzzy num-
bers? a numerical study. International Journal of Approximate Reasoning,
54(5):627 – 639, 2013.
6. S. M. Chen. New methods for subjective mental workload assessment and
fuzzy risk analysis. Cybernetics and Systems, 27(5):449–472, 1996.
7. S.-M. Chen and K. Sanguansat. Analyzing fuzzy risk based on a new fuzzy
ranking method between generalized fuzzy numbers. Expert Systems with
Applications, 38(3):2163 – 2171, 2011.
8. Z. Chen, G. Huang, and A. Chakma. Hybrid fuzzy-stochastic modeling
approach for assessing environmental risks at contaminated groundwater
systems. Journal of Environmental Engineering, 129(1):79–88, 2003.
9. F. Choobineh and H. Li. An index for ordering fuzzy numbers. Fuzzy Sets
and Systems, 54(3):287 – 294, 1993.
10. R. Chutia. Ranking of fuzzy numbers by using value and angle in the epsi-
lon-deviation degree method. Applied Soft Computing, 60:706 – 721, 2017.
11. R. Chutia and B. Chutia. A new method of ranking parametric form of fuzzy
numbers using value and ambiguity. Applied Soft Computing, 52:1154 – 1168,
2017.
12. R. Chutia, R. Gogoi, and D. Datta. Ranking p-norm generalised fuzzy
numbers with different left height and right height using integral values.
Mathematical Sciences, 9(1):1–9, 2015.
13. I. Deli. Operators on single valued trapezoidal neutrosophic numbers and
svtn-group decision making. Neutrosophic Sets and Systems, 22:131–150.
360 Mathematics and Computer Science Volume 2
14. I. Deli and Y. uba. A ranking method of single valued neutrosophic num-
bers and its applications to multi-attribute decision making problems.
International Journal of Machine Learning and Cybernetics, 8, 08 2017.
15. D. Dubois and H. Prade. Fuzzy Sets and Systems: Theory and Applications.
Academic Press, Inc., Orlando, FL, USA, 1980.
16. T. Garai, H. Garg, and T. Roy. A ranking method based on possibility mean
for multi-attribute decision making with single valued neutrosophic num-
bers. Journal of Ambient Intelligence and Humanized Computing, 11, 11 2020.
17. P. Grzegorzewski. Algorithms for trapezoidal approximations of fuzzy num-
bers preserving the expected interval. Fuzzy Sets and Systems, 159:1354–
1364, 06 2008.
18. R. Jain. Decisionmaking in the presence of fuzzy variables. IEEE Transactions
on Systems, Man and Cybernetics, SMC-6(10):698–703, Oct 1976.
19. R. Jain. A procedure for multiple-aspect decision making using fuzzy sets.
International Journal of Systems Science, 8(1):1–7, 1977.
20. C. Jana, G. Muhiuddin, and M. Pal. Multiple-attribute decision-making
problems based on svtnh methods. Journal of Ambient Intelligence and
Humanized Computing, 11, 09 2020.
21. C. Jana and M. Pal. A robust single-valued neutrosophic soft aggregation
operators in multi-criteria decision making. Symmetry, 11(1), 2019.
22. C. Jana, M. Pal, F. Karaaslan, and J. Wang. Trapezoidal neutrosophic aggrega-
tion operators and their application to the multi-attribute decision-making
process. Scientia Iranica, 27(3):1655–1673, 2020.
23. F. KARAASLAN. Gaussian single-valued neutrosophic numbers and its appli-
cation in multi- attribute decision making. Infinite Study, 2018.
24. J. Peng, J. Wang, H.-Y. Zhang, and X.-h. Chen. An outranking approach for
multi-criteria decision-making problems with simplified neutrosophic sets.
Applied Soft Computing, 25:336–346, 12 2014.
25. R. A. Shureshjani and M. Darehmiraki. A new parametric method for rank-
ing fuzzy numbers. Indagationes Mathematicae, 24(3):518 – 529, 2013.
26. F. Smarandache. Neutrosophy: Neutrosophic Probability, Set, and Logic:
Analytic Synthesis & Synthetic Analysis. Analytic Synthesis and Synthetic
Analysis Series. American Research Press, 1998.
27. F. Smarandache. A unifying field in logics: Neutrosophic logic., 1999.
28. F. Smarandache. Neutrosophic set a generalization of the intuitionistic fuzzy
set. International Journal of Pure and Applied Mathematics, 24:38–42, 01
2004.
29. V. Ulucay, A. Kilic, I. Yildiz, and M. ahin. A new approach for multi-attribute
decision-making problems in bipolar neutrosophic sets. Infinite Study, 2018.
30. X. Wang and E. E. Kerre. Reasonable properties for the ordering of fuzzy
quantities (i). Fuzzy Sets and Systems, 118(3):375 – 385, 2001.
31. X. Wang and E. E. Kerre. Reasonable properties for the ordering of fuzzy
quantities (ii). Fuzzy Sets and Systems, 118(3):387 – 405, 2001.
Ranking of Single Valued Neutrosophic Fuzzy Numbers 361
32. V. F. Yu and L. Q. Dat. An improved ranking method for fuzzy numbers with
integral values. Applied Soft Computing, 14, Part C:603 – 608, 2014.
33. L. A. Zadeh. Fuzzy sets. Information and Control, 8(3):338–353, 1965.
34. H.-Y. Zhang, J. Wang, and X. Chen. An outranking approach for multi-
criteria decision-making problems with interval-valued neutrosophic sets.
Neural Computing and Applications, 27, 04 2015.
25
Performance Analysis of Database Models
Based on Fuzzy and Vague Sets for
Uncertain Query Processing
Sharmistha Ghosh1* and Surath Roy2
Abstract
One of the primary aspects of utilization of any database model lies in its potential
in processing information and queries accurately. In the present work, the authors
intend to make a comparative analysis on the capability of fuzzy and vague rela-
tional database models in treating uncertain queries. A new algorithm is proposed
and query testing related to a real life example is performed. The investigation
demonstrates that a relational data model based on vague set theory produces
more refined decisions than a fuzzy data model. It may thus be asserted that a
relational database management system (RDBMS) using vague theoretic concept
might lead to better software fabrication than the presently accessible ones.
Keywords: Fuzzy set, vague set, database model, similarity measure, SQL,
fuzzy SQL, vague SQL
25.1 Introduction
In real life, information is very often imprecise or incomplete. The conven-
tional relational database system fails to treat such uncertain data. The the-
ory of fuzzy sets, as formalized by Zadeh [20] in 1965, is extensively applied
to handle such inexact or imprecise data. Several authors [1, 6, 10, 11, 14, 15,
17–19] have also worked on formulation of a query language for a database
representation based on fuzzy sets. Bosc et al. [1] and Nakajima et al. [15]
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (363–384) © 2023 Scrivener Publishing LLC
363
364 Mathematics and Computer Science Volume 2
have outstretched the familiar SQL language in the framework of fuzzy set
theory, namely, SQLF. More recently, Moreau et al. [14] have discussed a
procedure wherein one may fetch data with no prior information about the
database constitution or precise query language. The potential benefits and
efficacy of the use of fuzzy query in a traditional data model have been thor-
oughly explained in [17]. An intelligent approach to extend SQL language to
treat flexible conditions in queries was proposed by Mama et al. [11] in 2021.
Gau and Buehrer [4] initiated the theory of vague sets in 1993 as an
abstraction of a fuzzy set theoretic approach. It is believed that a vague set
that employs interval-based membership values may process uncertain
information in a more efficient manner compared to a fuzzy set. The vague
theoretic concept was further embodied in relations by Lu and Ng [8, 9] and
a new query language called VSQL evolved. A vague database model was
designed by Zhao and Ma [21] and vague querying strategies with SQL were
also investigated. In [3], Dutta et al. have used vague sets to rejuvenate “vague
search”, a new method of intelligent search that is capable of answering any
uncertain query put forth by the user. In [12], the detailed design aspects of
a vague database model have been thoroughly discussed. An architecture for
processing of hesitant queries has been designed by Mishra et al. [13] along
with a comparison of fuzzy and vague sets in handling imprecise queries.
In the present work, the authors also aspire to analyze the ability of fuzzy
and vague database models in relation to uncertain query processing. It has
been noticed that the algorithms used in literature (see refs. [10, 13, 16]) to
obtain membership values possess certain drawbacks. In the current inves-
tigation, the authors present a novel algorithm which is devoid of such defi-
ciencies. The proposed algorithm generates membership functions for fuzzy
or vague sets for the calculation of membership values that are independent
of the attribute. However, it is worthwhile to mention that the membership
functions employed earlier in literature [10, 16] depend on the attribute type.
The framework of the chapter is as follows. The definitions of fuzzy
and vague sets as well as related fundamental concepts are presented in
Sections 25.2.1 and 25.2.2. The similarity measure formulae used in this
study appear in Section 25.2.3. The algorithm designed in the present anal-
ysis to generate membership values is proposed in Section 25.3. Real life
examples are demonstrated in Section 25.4 to observe that a vague theo-
retic approach is better suited in processing queries that are not precise.
The concluding remarks are reported in Section 25.5.
1 if x ∈A
µA =
0 if x ∉A
shown to be more effective in general cases. This measure has been used in
this work, which is presented below:
(t1 − t2 ) − ( f1 − f2 )
SM (u,=
v) 1 −
2
(
1 − ( t1 − t 2 ) + ( f1 − f 2 ) . )
Definition 25.2.3.2: Similarity Measure for Vague Sets
Let X = {x1 , x2 , x3, ..., xn } be the universe. Let S1 and S2 be two vague sets
of X, such that
=S1 { x ,[t
i S1 (xi ),1 − f S1 (xi )] : xi ∈ X} ,
where t S1 ( xi ) ≤ µ S1 ( xi ) ≤ 1 − f S1 ( xi ) and 1 ≤ i ≤ n
and
S2 = { xi ,[t S2 ( xi ),1 − f S2 ( xi )] : xi ∈ X } ,
where t S2 ( xi ) ≤ µ S2 ( xi ) ≤ 1 − f S2 ( xi ) and 1 ≤ i ≤ n.
Now, the similarity measure between S1 and S2, denoted by SM (S1 , S2 ) ,
is defined as:
SM ( S1 , S2 ) =
1
n ∑ SM ([t
i =1
S1 ( xi ),1 − f S1 ( xi )],[t S2 ( xi ),1 − f S2 ( xi )])
1
µclose to u ( x ) = .
|x − u|
1+
20000
However, for the Experience attribute, Raju et al. defined the member-
ship function as
1
µclose to u ( x ) = .
|x − u|
1+
4
1
µclose to u ( x ) = 2 ;
|x − u|
1+
20000
1
µclose to u ( x ) = 2.
|x − u|
1+
4
Algorithm 25.3.1
Input: Fuzzy or vague data and attributes in an imprecise query
Output: Membership value lying in [0,1]
Method:
Fetch attributes from the query
for each fuzzy or vague attribute do
begin
f_data = data value of the attribute from the query
range = maximum domain value – minimum domain value
for every tuple in the relation do
begin
t_value = tuple value from the domain of fuzzy or vague attribute
if (|f_data – t_value | ≤ range)
|f _ data − t _ value|
membership _ value = 1 −
range
else membership_value = 0
end for loop for tuple
end for loop for fuzzy or vague attribute
The fuzzy and vague database models were used to execute the above
queries with one or more attributes. In each case, the vague database model
produced better results than the fuzzy data model, as may be observed
from the analysis presented herein.
t1 = 1, f1 = 0, t2 = 0.28125, f2 = 0.71875
Mr. Bose 57 <57, 0.875> <57, [0.875, 0.875]> 0.93541 22 77900 0.93541
Mr. Ghosh 49 <49, 0.875> <49, [0.875, 0.875]> 0.93541 17 57400 0.93541
Mr. Bhowmick 41 <41, 0.625> <41, [0.625,. 0.625]> 0.79057 13 46500 0.79057
Mr. Banerjee 39 <39, 0.5625 > <39, [0.5625, 0.75000 12 42000 0.75000
0.5625]>
Table 25.3 Similarity measures for Query 1 with Fuzzy Model. (Continued)
Fuzzy Vague
Age representation representation SM of age with Experience Remuneration
Name (years) of age of age <53,[1,1]> (years) (₹) SM (tuple)
Mr. Ganguly 52 <52, 0.96875> <52, [0.96875, 0.98425 19 80200 0.98425
0.96875]>
Mr. Kher 56 <56, 0.90625 > <56, [0.90625, 0.95197 21 81100 0.95197
0.90625]>
Mr. Mukhopadhyay 54 <54, 0.96875 > <54, [0.96875, 0.9842 23 82200 0.9842
0.96875]>
and thus,
SM (u, v )
(1 − 0.28125) − (0 − 0.71875)
= 1 − (1 − (1 − 0.28125) + (0 − 0..71875) )
2
= 1 −=
0.71875 =
0.28125 0.53033
t1 = 1, f1 = 0, t2 = 0.3125, f2 = 0.6875.
Then,
SM (u, v )
(1 − 0.3125) − (0 − 0.6875)
= 1 − (1 − (1 − 0.3125) + (0 − 0.68775) )
2
= 1 −=
0.6875 =
0.3125 0.55902
Table 25.4 Resulting tuples for Query 1 with Fuzzy Model at α= 0.95.
Name Age (years) Experience (years) Remuneration (₹)
Mr. Barik 51 18 58000
Mr. Samanta 50 16 56800
Mr. Chattopadhyay 53 20 80900
Mr. Ganguly 52 19 80200
Mr. Kher 56 21 81100
Mr. Mukhopadhyay 54 23 82200
Mr. Patnaik 55 27 82000
374 Mathematics and Computer Science Volume 2
Query 2: “To retrieve the data of the employees whose age is more
or less than 53 years with work experience close to 22 years.”
Table 25.6 Resulting tuples for Query 1 with Vague Model at α = 0.95.
Name Age (years) Experience (years) Remuneration (₹)
Mr. Barik 51 18 58000
Mr. Samanta 50 16 56800
Mr. Chattopadhyay 53 20 80900
Mr. Ganguly 52 19 80200
Mr. Mukhopadhyay 54 23 82200
Mr. Patnaik 55 27 82000
376 Mathematics and Computer Science Volume 2
Mr. Roy 30 <30, 0.28125> <30, [0.28125, 0.53033 4 <4,0.419355> <4,[0.4193550 0.647576 32000 0.53033
0.28125]> .419355]>
Mr. Roychowdhury 31 <31, 0.3125> <31, [0.3125, 0.55902 6 <6,0.483871> <6,[0.483871, 0.695608 32500 0.55902
0.3125]> 0.483871]>
Mr. Pramanik 34 <34, 0.40625> <34, [0.40625, 0.63738 8 <8, 0.548387> <8,[ 0.548387, 0.740532 34000 0.63738
0.40625]> 0.548387]>
Mr. Barik 51 <51, 0.9375> <51, [0.9375, 0.96825 18 <18,.0.870968> <18,[0.870968, 0.933257 58000 0.933257
0.9375]> 0.870968]>
Mr. Bose 57 <57, 0.875> <57, [0.875, 0.93541 22 <22,1> <22,[1,1]> 1.000000 77900 0.93541
0.875]>
Mr. Ghosh 49 <49, 0.875> <49, [0.875,. 0.93541 17 <17,0.838710> <17,[0.838710, 0.915811 57400 0.915811
0.875]> 0.838710]>
Mr. Samanta 50 <50, 0.90625> <50, [0.90625, 0.95197 16 <16,0.806452> <16,[0.806452, 0.898027 56800 0.898027
0.90625]> 0.806452]>
Mr. Bhowmick 41 <41, 0.625> <41, [0.625,. 0.79057 13 <13,0.709677> <13,[0.709677, 0.842424 46500 0.79057
0.625]> 0.709677]>
Mr. Banerjee 39 <39, 0.5625 > <39, [0.5625, 0.75000 12 <12,0.677419> <12,[0.677419, 0.823055 42000 0.75000
0.5625]> 0.677419]>
Mr. Chattopadhyay 53 <53, 1> <53, [1,1]> 1 20 <20,0.935484> <20,[0.935484, 0.967204 80900 0.967204
0.935484]>
(Continued)
Performance Analysis of Database Models 377
Table 25.7 Similarity Measures for Query 2 with Fuzzy Model. (Continued)
Vague Fuzzy SM of
Fuzzy represen represen SM of age represen Vague represen experience
Age tation of tation of with <53, Experience tation of tation of with Remuneration SM
Name (years) age age [1,1]> (years) experience experience <22,1,1]> (₹) (tuple)
Mr. Ganguly 52 <52, 0.96875> <52, [0.96875, 0.98425 19 <19,.0.903226> <19,[.0.903226, 0.950382 80200 0.950382
0.96875]> .0.903226>
Mr. Kher 56 <56, 0.90625 > <56, [0.90625, 0.95197 21 <21,0.967742> <21,[0.967742, 0.983739 81100 0.95197
0.90625]> 0.967742]>
Mr. Mukhopadhyay 54 <54, 0.96875 > <54, [0.96875, 0.9842 23 <23,0.967742> <23,[0.967742, 0.983739 82200 0.983739
0.96875]> 0.967742]>
Mr. Patnaik 55 <55, 0.9375> <55, [0.9375, 0.96825 27 <27,0.838710> <27,[0.838710, 0.915811 82000 0.915811
0.9375]> 0.838710]>
Mr. Sahoo 62 <62, 0.71875> <62, [0.71875, 0.84779 35 <35,0.580645> <35,[0.580645, 0.762001 122000 0.762001
0.71875]> 0.580645]>
378 Mathematics and Computer Science Volume 2
Table 25.8 Resulting tuples for Query 2 with Fuzzy Model at α = 0.95.
Age Experience Remuneration
Name (years) (years) (₹)
Mr. Chattopadhyay 53 20 80900
Mr. Ganguly 52 19 80200
Mr. Kher 56 21 81100
Mr. Mukhopadhyay 54 23 82200
Table 25.9 Resulting tuples for Query 2 with Fuzzy Model at α = 0.91.
Age Experience Remuneration
Name (years) (years) (₹)
Mr. Barik 51 18 58000
Mr. Bose 57 22 77900
Mr. Ghosh 49 17 57400
Mr. Chattopadhyay 53 20 80900
Mr. Ganguly 52 19 80200
Mr. Kher 56 21 81100
Mr. Mukhopadhyay 54 23 82200
Mr. Patnaik 55 27 82000
Mr. Roychowdhury 31 <31, [0.3125, 0.485216]> 0.574358 6 <6,[0.483871, 0.695474 32500 0.574358
0.560575]>
Mr. Pramanik 34 <34, [0.40625, 0.640691 8 <8,[ 0.548387, 0.739961 34000 0.640691
0.516927]> 0.603663]>
Mr. Barik 51 <51, [0.9375, 0.953667]> 0.964182 18 <18,[0.870968, 0.931279 58000 0.931279
0.888225]]>
Mr. Bose 57 <57, [0.875, 0.878291]> 0.929415 22 <22,[1,1]> 1.000000 77900 0.929415
Mr. Ghosh 49 <49, [0.875, 0.875770]> 0.930180 17 <17,[0.83871, 0.910255 57400 0.910255
0.884063]>
(Continued)
380 Mathematics and Computer Science Volume 2
Table 25.10 Similarity measures for Query 2 with Vague Model. (Continued)
SM of age Vague SM of experience
Age Vague representation with <53, Experience representation with <22, Remuneration SM
Name (years) of age [1,1]> (years) of experience [1,1]> (₹) (tuple)
Mr. Chattopadhyay 53 <53, [1.000000, 1.000000 20 <20,[0.935484, 0.963133 80900 0.963133
1.00000]> 0.942255]>
Mr. Ganguly 52 <52, [0.96875, 0.982317 19 <19,[.0.90326, 0.945563 80200 0.945563
0.977372]> 0.914820>
Mr. Kher 56 <56, [0.90625, 0.946656 21 <21,[0.967742, 0.983458 81100 0.946656
0.914506]> 0.975348]>
Mr. 54 <54, [0.96875, 0.982701 23 <23,[0.96774, 0.982917 82200 0.982701
Mukhopadhyay 0.976538]> 0.971200]>
Mr. Patnaik 55 <55, [0.9375, 0.943755]> 0.966286 27 <27,[0.83871, 0.908944 82000 0.908944
0.873021]>
Mr. Sahoo 62 <62, [0.71875, 0.835870 35 <35,[0.58064, 0.756300 122000 0.756300
0.783282]> 0.667873]>
Performance Analysis of Database Models 381
Table 25.11 Resulting tuples for Query 2 with Vague Model at α = 0.95.
Age Experience Remuneration
Name (years) (years) (₹)
Mr. Chattopadhyay 53 20 80900
Mr. Mukhopadhyay 54 23 82200
Table 25.12 Resulting tuples for Query 2 with Vague Model at α = 0.91.
Age Experience Remuneration
Name (years) (years) (₹)
Mr. Barik 51 18 58000
Mr. Bose 57 22 77900
Mr. Ghosh 49 17 57400
Mr. Chattopadhyay 53 20 80900
Mr. Ganguly 52 19 80200
Mr. Kher 56 21 81100
Mr. Mukhopadhyay 54 23 82200
382 Mathematics and Computer Science Volume 2
25.5 Conclusion
An algorithm has been designed in this research work to generate the fuzzy
or vague representation of attributes for an imprecise query. The suggested
technique produces a membership function for calculating membership
values that do not depend on the type of the attribute and are free of vari-
ous drawbacks identified in existing membership functions studied in the
literature. Using an Employee database, a comparison of fuzzy and vague
sets for handling imprecise queries has been performed. The results of
this study confirms that a database model based on vague sets can process
uncertain queries more efficiently than a fuzzy model.
References
1. Bosc, P. and Pivert, O. (1995). SQLF: A relational database language for fuzzy
querying. IEEE Transaction on Fuzzy Systems, Vol. 3, No. 1, pp. 1-17.
2. Chen, S. M. (1997). Similarity Measure between Vague Sets and between
Elements. IEEE Trans. Systems. Man and Cybernetics, Vol. 27, No. 1,
pp. 153-158.
3. Dutta, A. K., Idwan, S. and Biswas, R. (2009). A Study of Vague Search to
Answer Imprecise Query. International Journal of Computational Cognition,
Vol. 7, No. 4, pp. 63-69.
4. Gau, W. L. and Buehrer, D. J. (1993). Vague Sets. IEEE Trans. Syst. Man,
Cybernetics, Vol. 23, No. 2, pp. 610-614.
5. Hong, D. H. and Kim, C. (1999). A Note on Similarity Measures between
Vague Sets and between Elements. Information Sciences, Vol. 115, pp. 83-96.
6. Intan, R. and Mukaidono, M. (2000). Fuzzy functional dependency and its
application to approximate data querying. In Proceedings of International
Database Engineering and Applications Symposium, pp. 47-54.
7. Li, F. and Xu, Z. (2001). Measures of Similarity between Vague Sets. Journal
of Software, Vol. 12, No. 6, pp. 922-927.
8. Lu, A. and Ng, W. (2004). Managing Merged Data by Vague Functional
Dependencies, Lecture Notes in Computer Science, Vol. 3288, pp. 259-272.
9. Lu, A. and Ng, W. (2005). Vague Sets or Intuitionist Fuzzy Sets for Handling
Vague data: Which One Is Better? Lecture Notes in Computer Science, Vol.
3716, pp. 401-416.
10. Ma, Z. M. and Meng, X. (2008). A Knowledge-Based Approach for
Answering Fuzzy Queries over Relational Databases, Lecture Notes in
Artificial Intelligence, Vol. 5178, pp. 623-630.
Performance Analysis of Database Models 383
11. Mama, R. and Machkour, M. (2021). Fuzzy querying with SQL: Fuzzy view-
based approach. Journal of Intelligent and Fuzzy Systems, Vol. 40, No. 5,
pp. 9937-9948.
12. Mishra, J. (2014). An Extension of Fuzzy Relational Database Model into
Vague Relational Database Model, Ph.D. Thesis, West Bengal University of
Technology.
13. Mishra, J. and Ghosh, S. (2014). Uncertain Query Processing using Vague
Set or Fuzzy Set: Which One Is Better? International Journal of Computers
Communications & Control, Vol. 9, No. 6, pp. 730-740.
14. Moreau, A., Pivert, O. and Smits, G. (2018). Fuzzy Query by example, 33rd
ACM/SIGAPP Symposium on Applied Computing (SAC-2018), France.
15. Nakajima, H., Sogoh, T. and Arao, M. (1993). Fuzzy Database Language and
Library- Fuzzy Extension to SQL. Proceedings of Second IEEE International
Conference on Fuzzy Systems, Vol. 1, pp. 477-482.
16. Raju, K.V.S.V.N. and Majumdar, A. K. (1988). Fuzzy functional dependen-
cies and lossless join decomposition of fuzzy relational database system,
ACM Transactions on Database Systems, Vol. 13, No. 2, pp. 129-166.
17. Srivastava, A., Yadav, S., Srivastava, N. and Khan, Z. (2016). Fuzzy Query: An
Impression in Query processing, Proceedings of IEEE Sponsored International
Conference on Advancement in Computer Engineering and Information
Technology (ACEIT 2016).
18. Takahashi, Y. (1993). Fuzzy database query languages and their rela-
tional completeness theorem, IEEE Transactions on Knowledge and Data
Engineering, Vol. 5, pp. 122-125.
19. Yadav, R. S. (2019). A Study of SQL query processing using soft comput-
ing techniques: a hybrid vague logic approach, International Journal of
Information Technology, Vol. 11, pp. 393-405.
20. Zadeh, L. A. (1965). Fuzzy Sets, Information and Control, Vol. 8, No. 3,
pp. 338-353.
21. Zhao, F.and Ma, Z. M. (2009). Vague Query Based on Vague Relational
Model, Advances in Computational Intelligence - Part of AINSC Book Series,
Vol. 116, pp. 229-238.
26
Estimating Error of Signals by Product
Means (N , pn , qn) (C, 2) of the Fourier
Series in a W(Lr, ξ (t))(r ≥ 1) Class
Pankaj Tiwari* and Aradhana Dutt Jauhari†
Abstract
The goal of this paper is to approach a new result on the degree of Fourier series
r
approximation of a function g ∈ W(L , ξ(t))(r ≥ 1) class by (N , pn , qn ) (C, 2) prod-
uct summability. Signals are handled as one-variable functions, while images are
represented as two-variable functions. The concept’s research is directly tied to the
rapidly developing field of information technology. The approximation theory is
a trigonometric polynomial approximation. Many academics have researched in
similar lines, but our research also proves some fresh findings.
26.1 Introduction
Let L = L(0, 2π) denote the space of 2π- periodic and Lebesgue integrable
functions on the (0, 2π) Fourier series at point x:
∞
a
g 0 +
2 ∑(a + cosnx + b sinnx)
n =1
n n
Let ∑an be an nth partial sum sequence {sn} in an infinite series [9].
Sharmistha Ghosh, M. Niranjanamurthy, Krishanu Deyasi, Biswadip Basu Mallik, and Santanu Das (eds.)
Mathematics and Computer Science Volume 2, (385–396) © 2023 Scrivener Publishing LLC
385
386 Mathematics and Computer Science Volume 2
26.1.1 Definition 1
The transformation from sequence to sequence [5]:
N 1 n
t=
n ∑
= k 0 pn−k qk sk (26.1)
Rn
26.1.2 Definition 2
Riesz means can also be expressed as:
1 n
tnN == ∑ k =0 pkqk sk (26.2)
Rn
26.1.3 Definition 3
The (C,2) summability nth partial sum of the (C,2) transform is given [3]-.
2
Cn2 = ∑nk =0 (n − k + 1)sk → s as n → ∞ (26.3)
(n + 1)(n + 2)
By using the (C,2) method, the infinite series ∑an can be summable to
the definite as the number.
Estimating Error of Signals 387
26.1.4 Definition 4
(N, pn , q n ) of {sn} is a composite transformation over (C,2) as
∑
1 n
TnNC = pkqkCk2
Rn k =0
(26.4)
2 n pkqk (n − k + 1)
TnNC
= ∑ k =0 sk
Rn (k + 1)(k + 2)
26.1.5 Remark 1
If qn = 1, then in (26.2) the (N , pn , qn ) summability method reduces with
pn = 1(N , pn ) summability and it reduces to (N , qn ) summability.
Also, the degree to which of the function g: R → R is approximated by a
trigonometric polynomial tn of degree n:
1/r
2π
∫ O(|t|α ), 0 < α ≤ 1, r ≥ 1, t > 0 (26.7)
r
|( g (x + t ) − g (x ))
= | dx
0
(∫ )
2π 1/r
0 )) |r dx
|( g (x + t ) − g (x= O(ξ (t )), r ≥ 1,t > 0 (26.8)
388 Mathematics and Computer Science Volume 2
(∫ )
2π 1/r
β
0 |( g (x + t ) − g (x ))}sin= x |r dx O(ξ (t )), β ≥ 0 ,t > 0. (26.9)
ψ=
(t ) g (x + t ) + g (x − t ) − 2 g (x )
1 p q (n − k + 1) sin2 (n − k + 2)t /2
=
Pn (t ) ∑nk =0 k k sk (26.10)
2π Rn (k + 1)(k + 2) sin2t /2
r
Then, the weighted W(L , ξ(t)) class is a subset of the Lipα, Lip(α, r), and
Lip(ξ(t), r) classes. As a result, the following additions have been made:
26.2.1 Theorem 1
If f ∈ W(Lr, ξ(t)) class is a 2π-periodic function that is integrable in the
Lebesgue sense in [0,2π], then the degree of approximation is given by
β+
1
1
Tn
NE
O (n + 1) r ξ
− gr =
(n + 1)
where TnNE is the (N , pn , qn ) (E,s) transform of {sn}, provided ξ(t) has the
following requirements: { }
ξ(t )
t
is a decreasing sequence,
Estimating Error of Signals 389
1 r 1/r
(n +1)
t |φ (t )| 1
∫ 0
ξ (t )
sin βr
t dt = O
n + 1
r 1/r
π
t −δ |φ (t )|
∫ 1
(n +1)
ξ (t ) dt = O(n + 1)
δ
26.3.1 Theorem 2
1
If g ∈ W(Lr, ξ(t)) with 0 ≤ β ≤ 1 − , the degree of approximation by Riesz-
r
Cesaro product mean of the Fourier series satisfies for n = 0,1,2…
1 β+
1
TnNC − g (x )r =
O ξ n(n + 2) r
(26.11)
(n + 2)
{ }
ξ (t )
t
is a function that does not increase and
1 r 1/r
(n + 2 ) t |ψ (t )sin β t| n
∫ 0
ξ (t )
dt = O
n + 2
(26.12)
r 1/r
π
t −δ |ψ (t )| sin β t
∫ 1
(n + 2 )
ξ (t )
dt = O(n + 2)δ (26.13)
26.4.1 Lemma-1
For
1 1 π π
0≤t ≤ , ≤ for 0 < t < ,
(n + 2) sint 2t 2
n
pk qk (n − k + 1) sin2 (n − k + 2)t /2
Proof : |Pn (t )| =
1
2π Rn ∑
k =0
(k + 1)(k + 2)
sk
sin2t /2
1 n sk (n − k + 1) (n − k + 2)t 2 /π 2
≤ ∑ k =0
2π (k + 1)(k + 2) t 2 /π 2
= O (n+ 2)
26.4.2 Lemma-2
For (n + 2)−1 < t < π,
n
we get Pn (t ) = O( )
(n + 2)t 2
1 n pk qk (n − k + 1) sin 2 (n − k + 2)t /2
Proof : |Pn (t )| = ∑ k =0 sk
2π Rn (k + 1)(k + 2) sin 2t /2
t t
Using Jordan’s Lemma sin ≥ and sinkt ≤, we have
2 π
1 π 2 n pk qk (n − k + 1)
≤ ∑ k =0 sk
2π Rn t 2 (k + 1)(k + 2)
n
= O
(n + 2)t 2
Estimating Error of Signals 391
π
1 sin 2 (n − k + 2)t /2
sn ( g ) − g ( x ) =
2π ∫ 0
ψ (t )
sin 2t /2
dt
1 p q (n − k + 1) π sin2 (n − k + 2)t /2
{TnNC − g (x )} =∑nk =0 k k sk ∫ 0 ψ (t ) dt
2π Rn (k + 1)(k + 2) sin2t /2
π
=
∫ ψ (t )P (t )dt
0
1
n
(n+2 ) π
= ∫0 ψ (t )Pn (t )dt + ∫ 1 ψ (t )Pn (t )dt
(n+2 )
= I1 + I 2 (26.14)
1
( n+ 2)
where |I1| ≤ ∫ 0 |ψ (t )||Pn (t )| dt
Applying Minikowski’s inequality, since
2π 1/r
∫ 0
|ψ ( x + t ) − ψ ( x )sin x| dx
β r
2π 1/r
≤
∫ 0
|g (u + x + t ) − g (u + x )sin x| dx
β r
1/r
2π
+
0∫ |g (u − x − t ) − g (u − x )sin β x|r dx
= O{ξ (t )}
{ }
I /r 1/s
1
t |ψ (t )| sin β t r 1
ξ (t )| Pn (t ) s
∫ ∫
(n + 2) (n + 2)
I1 ≤ dt β dt
0 ξ (t ) 0 tsin t
{ }
1 1/ s
n ξ (t )|Pn (t )| s
= O
∫
(n + 2)
dt {using (26.12)}
(n + 2) 0 tsin t
β
{ }
1 1/ s
n
s
ξ (t )O(n + 2)
=
∫
(n + 2)
dt {using Lemma 1}
(n + 2) 0 tsin β t
Using the second mean value theorem for integrals, ξ(t) is a decreasing
function and we get
1 1/s
1 dt
1
≤ O nξ
∫
(n + 2)
0 ≤ε <
(n + 2) 0 t ( β +1)s (n + 2)
1 1/s
1
∫ [t −( β +1)s dt ]
(n + 2)
= O nξ
(n + 2) 0
1/(n + 2) 1/s
1 [t
− ( β +1)s +1
]
= O nξ
(n + 2) −(1 + β )s +1 ε
− β −1/ r
1 1
= O nξ
(n + 2) (n + 2)
1
= O ξ n(n + 2)β +1/r (26.15)
(n + 2)
Now,
π
|I 2 | ≤ ∫ 1 |ψ (t )||Pn (t )|dt
(n + 2)
2t
Using Hölder’s inequality as a starting point, ≤ |sint| < 1 [10]
π
Estimating Error of Signals 393
1/r 1/s
t −δ |ψ (t )|sin β t
r
ξ (t )|Pn (t )|
s
π π
≤
∫
1/(n+2) ξ (t )
dt ⋅
∫ −δ β dt
1/(n+2) t sin t
1/s
n
s
π ξ (t )O(
δ (n + 2)
= O(n + 2)
∫ t −δ + β +2 dt using (26.4)
1/(n+2)
{ }
1/s
π
ξ (t )
s
= O(n + 2)δ −1n
∫1/(n+2) t −δ + β + 2
dt by Lemma (2)
1/ s
1
s
(n + 2) ξ
δ −1 x dx
= O{n(n + 2) }
∫ 1/π
x δ − β − 2 x 2
1
ξ
x
Because ξ(t) > 0 and are both rising functions, the second mean
(1 / x )
value theorem is used and
1/ s
1 (n + 2)
dx
= O n(n + 2)δ −1 ξ
(n + 2) ∫1/π x δ s − β s −2s+2
1/ s
δ −1 1 x β s + 2 s −δ s − 2+1
= O n(n + 2) ξ
(n + 2) β s + 2s − δ s − 1
(n + 2)
1 x ( β −δ + 2)−1 s
= O n(n + 2)δ −1 ξ
(n + 2) ( β − δ + 2) − 1
s 1
π
1
= O ξ n(n + 2)β +1/r (26.16)
(n + 2)
394 Mathematics and Computer Science Volume 2
1
TnNC − g (x ) r =
O ξ n(n + 2)β +1/ r
(n + 2)
26.6 Applications
From our main theorem, we can derive the following corollaries.
26.6.1 Cor. 1
If ξ(t) = tα, then weighted class W ( Lr , ξ (t ) ) , 1 ≤ r < ∞ reduces to the group
of class Lip(α, r) and then a function f is a degree of approximation belong-
ing to the class Lip(α, r), r−1 ≤ α ≤ 1, given by
TnNC − g (x )r =+
O{n(n 2)1/ r }
26.6.2 Cor. 2
For r → ∞ in Cor. 1 for the class function, g ∈ Lip(α, r) reduces to the class
Lip α and a function g belonging to the class Lipα, 0 < α < 1 whose degree
of the approximation is given by
O{n(n 2)α }
TnNC − g (x )r =+
26.7 Conclusion
The previous finding on function degree of approximation demonstrated
that the result is frequently true in nature and can be simplified to a few
special cases. As a result, the current research can be used for a series of
issues in the fields of analysis, technology, and engineering. Figure 26.1
shows some unique behaviour in the Fourier series approximations [11].
Estimating Error of Signals 395
1 n=1
0 t
−1
1
n=5
0 t
−1
1 n=11
0 t
−1
1 n=49
0 t
−1
Figure 26.1 Approximations of such Fourier Series to sq (t). Each plot indicates the
number of terms in the Fourier sum and the square wave is displayed as a dashed line over
two periods.
Acknowledgement
The sincere gratitude goes to the anonymous referees for their careful
reading, remarks, and valuable comments, as well as many other useful
suggestions for improved presentation. The author wishes to express his
appreciation to the members of the editorial board.
References
1. H. K. Nigam: “Degree of approximation of a function belonging to weighted
(Lip (ξ(t), r) class by (C,1) (E,q) means”, Tamkang Journal of Mathematics, 42,
31-37 (2011).
2. H. K. Nigam: “On Approximation of functions by product operators”, Survey
in Mathematics and its Applications ISSN 1842-6298(Electronic), 1843-
7265(print) Volume 8 (2013), 125-136.
3. K. Sharma. and S.S. Malik: “Degree of Approximation by Product means
of the Fourier Series in a (W ( Lr , ξ(t ) ) (r > 1) − class by (E,1)(C,2) product
summability Transform”, International Journal of Mathematical Trends
and Technology (IJMTT), Volume-8 Number-2,ISSN 2231-5373,2014,
10.14445/22315373/IJMTT-V8P518.
396 Mathematics and Computer Science Volume 2
397
398 About the Editors
Best Research journal reviewer award twice. He had also received ear-
lier Young Researcher award in Computer Science and Engineering. He
obtained Patent Award - 2021 from KSCST, Karnataka, India. He con-
ducted various National and International Conferences, National Level
workshops and delivered lectures. Dr. Niranjanamurthy is associated
with various professional bodies as follows: IEEE Member, Life Member
of “The International Association of Engineers” (IAENG), and Member
of “Computer Science Teachers Association” (CSTA). His areas of interest
are Data Science, ML, E-Commerce, and M-Commerce related to Industry
Internal tool enhancement, Software Testing, Software Engineering, Web
Services, Web-Technologies, Cloud Computing, Big Data Analytics, and
Networking.
401
402 Index
Density of the fluid, 224 Elastic parameter, 117, 224, 231, 235,
Despite, 85, 97, 139, 206, 207 236
Detection, 1, 11, 12, 13, 14, 20, 22, 30, E-Learning, 137, 138, 139, 140, 141,
33, 49, 51, 54, 59, 60, 61, 62, 63, 142
64, 67, 77 Electrical conductivity, 224
Differential equations, 103, 228, 256, Electronic, 80, 139, 141, 209, 251, 252
257, 259 Encoder-Decoder, 5, 91, 92, 94, 95, 96
Digital, 2, 55, 82, 83, 85, 88, 139, 141, Energy ratios, 126, 131, 134
142, 209, 248 Entropy, 49, 50, 51, 52, 53, 54, 96, 224,
Digital Transformation, 139 226, 286, 287
Dimensionless elastic parameter, 224 Environment, 11, 21, 28, 62, 79, 140,
Dimensions, 240, 267, 273 142, 152, 173, 175, 183, 184, 205,
Directives, 80 206, 207, 215, 218, 273, 349
Disaster, 80 Epilepsy disease, 187, 191, 192, 201,
Disease, 6, 34, 49, 50, 51, 52, 53, 54, 55, 203
85, 101, 102, 111, 187, 189, 191, EPILONT, 191, 192, 193, 194, 198, 199,
192, 198, 201, 202, 203, 207, 284 200, 201, 202
Distributed, 25, 81, 151, 154, 155, 166, EPSO, 191, 192, 193, 196, 198, 199,
169, 176, 208, 337 200, 201, 202
Diverse, 59, 84, 85, 154, 239, 284 Equipment, 141, 206
Doctors, 192, 284, 285 ESSO, 193, 194, 196, 198, 199, 200,
Domain knowledge, 188 201, 202
Dosages, 34, 36, 43, 44, 45 Evaluation, 1, 11, 52, 69, 116, 140, 147,
Douglas-Gunn, 42 187, 188, 189, 192, 203, 276, 286,
Doxorubicin, 33, 35, 36, 43, 44 343
Dynamic, 22, 28, 33, 34, 35, 36, 37, 41, EvoMSA, 62
45, 101, 102, 103, 106, 107, 111, Existence, 39, 116, 139, 218, 223, 305,
173, 174, 175, 176, 178, 183, 209, 306, 307, 314, 333, 349
226, 240, 241, 242, 327, 328, 329, Exploitation vector, 68
330, 331, 333, 344
Dynamic Model, 28, 34, 35 Facebook, 14, 85, 86, 138, 139
Dynamic network, 327, 329, 330 Face-To-Face Communication, 142
Dynamic viscosity, 224 Failed Machines, 337, 338, 340, 347
Dynamical systems, 34 Feasibility of flow, 331, 332, 333, 334
Fibonacci Search Method, 265, 268,
E-Business, 140 270, 271
E-Citizens, 139 Finite difference, 42
Education, 1, 14, 82, 83, 84, 139, 140, Flickr, 138, 139
141 Flow capacity, 327
EduSERE, 62 Fluid, 116, 117, 223, 224, 225, 226
Efficacy, 80, 84, 218, 364 F-measure, 59, 69, 70, 71, 72, 74
E-Government, 140 Focal, 193, 194, 196, 198, 202
E-Health, 140 Fokker-Planck, 33, 35, 38, 42
404 Index