Professional Documents
Culture Documents
c11 Rain Fall Prediction
c11 Rain Fall Prediction
1
Contents
o Abstract
o Introduction
o Existing System
o Disadvantages of existing system
o Proposed System
o Advantages of proposed system
o Algorithm
o Modules
o UML Diagrams
o Hardware requirements
o Software requirements
o Further Enhancement
o Conclusion
Abstract
The rainfall rate is one of the essential phenomena in the weather system, which has
a direct influence on the agriculture and biological sectors. This project aims to
develop a multiple linear regression model in order to predict the rate of
precipitation (PRCP), i.e., rainfall rate. It is based on some weather parameters, such
as temperature, wind speed, and dew point. A Python code using the Pytorch library
has been written to develop the model, which applies Artificial Neural Networks
and multiple linear Regressions in Machine Learning.
Introduction
o The process of predicting the state of the atmosphere for a specific location in the
future is called weather forecasting.
o Several methods are used to generate weather forecasting, each of which differ in
its accuracy and efficiency.
o The three important steps that precede the process of weather forecasting are to
collect atmospheric data as much as possible and to use it in numerical models to
predict the future state of the atmosphere.
o Applying machine learning techniques in weather forecasting can compensate
complex meteorological physics model.
o Using metrological data set, we chose supervised learning method, which is
multiple linear regression.
o The aim of our project is to develop a multiple linear regression model to predict
the rainfall rate correctly.
Existing system
o The Existing system uses back propagation neural network for rainfall prediction.
o Marker based motion Capture (MoCap) Framework algorithm is used.
o This model is being used by Xianggen Gan and he tested using the dataset from
1970 to 2000 which has 16 meteorological parameters.
o During the network training, the target error is set as 0.01 and learning rate is set
as 0.01.
o This model is also implemented on mat lab neural network.
Disadvantages
o It predicts accurate only for annual rainfall when compared with monthly rainfall
prediction.
o Here, the computers needs to perform few millions of calculations which are
expensive.
Proposed system
o We are applying machine learning techniques in weather forecasting, which can
compensate complex meteorological physics model.
o Using the metrological data set, we chose supervised learning method, which is
multiple linear regression, instead of unsupervised learning or reinforcement
learning.
o We used different regression types in machine learning, such as linear regression,
logistic, polynomial regression.
o We are using the simpler and most frequent method is linear regression for
prediction.
o This is done in four steps beginning with understanding the data set and
correlation between variables, deleting unwanted factors from the data set, dealing
with missing data and outliers, finally treating data to facilitate handling.
Advantages
o The Linear regression used here, is used to predict a numeric value (dependent
variable) form a set of features (predictors)
o Linear Regression
o Artificial Neural Networks
Modules
o User
o Admin
o Data Preprocessing
o Machine Learning Results
Modules Description
o User:
The User can register the first. While registering he required a valid user email
and mobile for further communications. Once the user register then admin can
activate the user. Once admin activated the user then user can login into our
system. User can upload the dataset based on our dataset column matched.
o Admin:
Admin can login with his login details. Admin can activate the registered
users. Once he activate then only the user can login into our system. Admin can
view the overall data in the browser.
Modules Description
o Multiple linear regression (MLR):
The Multiple Linear Regression (MLR) model to predict four weather
parameters which are (maximum and minimum temperature, relative humidity,
and the category of rainfall).They have applied two methods to predict rainfall
forecasting rainfall, which are Autocorrelation Function (ACF) and projected
error.
o The UML is a very important part of developing objects oriented software and
the software development process.
Use Case Diagram
The proposed system can be used to estimate the rainfall over the required
period so that the respective authorities can take precautions to prevent the
loss of life and property. The proposed system uses modified linear regression
approach to predict the rainfall that has less error percentage than compared
to most data mining techniques like clustering, back propagation which
provides the generalized values rather than estimate values. This data is used
to perform the necessary calculations to predict the rainfall from average
temperature and cloud cover of that particular district.
Conclusion
In this project, we have used multiple linear regression model to predict the rate of
precipitation (i.e., rainfall rate) based on some weather parameters taken as the
independent variables. The weather parameters are the mean temperature, maximum
temperature, minimum temperature, Dewpoint, sea level pressure, station pressure,
mean visibility and wind speed. The average of the mean square error between the
actual and predicted value during training and testing phase was calculated. It was
found that obtained results show that the mean square error between actual and
predicted values of the rainfall precipitation rate (PRCP) has been significantly
decreased during testing time. It has been found to be 85% when the amount of test
data equals the amount of training data, and 59% when more test data is used.
Explanation of this reduction needs supplementary research. for example, it may
indicate that the model used needs more data in the training phase.
Thank you!