Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 24

KIRTESH TIWARI

PGP - DATA SCIENCE AND BUSINESS ANALYTICS . PGPDSBA ONLINE SEP_C 2021


TIWARI.KIRTESH@GMAIL.COM

MRA PROJECT
MILESTONE 2
AGENDA

Suggestion of
Exploratory Analysis Use of Market Associations
Possible Combos
and Inferences Basket Analysis Identified 
with Lucrative Offers
PROBLEM STATEMENT

A Grocery Store shared Exploratory Analysis Use of Market Basket Associations Identified  A suggestion of
the transactional data Analysis (Association Possible Combos with
with you. Your job is to Rules) Lucrative Offers
identify the most
popular combos that
can be suggested to the
Grocery Store chain
Exploratory Analysis of data & Write Something about the Put the associations in a Write recommendations
after a thorough an executive summary (in PPT) association rules and their tabular manner Make discount offers or
analysis of the most of your top findings, supported relevance in this case Explain about support, combos (or buy two get one
commonly occurring by graphs. Add KNIME workflow Image  confidence, & lift values that free) based on the associations
sets of menu items in Are there trends across Write about threshold values are calculated      and your experience   
the customer orders. months/years/quarters/days of Support and Confidence
etc. that you are able to
The Store doesn’t have notice?
any combo meals. Can
you suggest the best
combo meals?
DATA HEAD

• Data has 20641 rows and 3 columns (Date, Order_id


and Product
NULL INFORMATION OF DATA

• There is no null value present in the data set


DATA SET INFO

• Data set has 2 object type and 1 int64 columns.


TYPE OF PRODUCT AND THEIR COUNT

• Poultry has highest number of order followed by Soda and


Cereals
PRODUCT WISE ORDER COUNT

• Poultry has highest number of


order followed by Soda and
Cereals
• Almost all the product has more
than 500 times orders
PRODUCT WISE ORDER TREND OVER YEARS
ORDER TREND -YEARLY

Products Order trend is in declining sharply from 2018 to


2020

https://public.tableau.com/app/profile/kirtesh.tiwari/
viz/MBA_16561519169540/
ProductsOrderedPerYearQuarterMonth_1#1
ORDER TREND –QUARTERLY

In 2018 product order increased from q1


to q3 but in 2019 orders are decreasing
Quarter on Quarter

https://public.tableau.com/app/profile/
kirtesh.tiwari/viz/
MBA_16561519169540/
ProductsOrderedPerYearQuarterMonth_1#
1
EDA SALES TREND -MONTHLY

https://
public.tableau.com/app/
profile/kirtesh.tiwari/viz/
MBA_16561519169540/
ProductsOrderedPerYearQu
arterMonth_1#1
PRODUCT WISE ORDER TREND

https://public.tableau.com/app/
profile/kirtesh.tiwari/viz/
MBA_16561519169540/
ProductsOrderedPerYearQuarterMon
th_1#1
MARKET BASKET ANALYSIS

• Use of Market Basket Analysis (Association Rules) -->Write Something about the association
rules and its relevance in this case -->Add KNIME workflow Image -->Write about threshold
values of Support and Confidence
ASSOCIATION RULES

• Association rule mining is a procedure which aims to observe frequently occurring patterns, correlations, or
associations from dataset /database. It is used to identify unique patterns and rules. Those patterns define
relationships and interactions between different items
• Association rule is dependent on two parameters :
1. Support: Support indicates how frequently the if/then relationship appears in the database.
2. Confidence: Confidence tells about the number of times these relationships have been found to be true.
• Lift – It is the ratio of the observed frequency of co-occurrence of our items and the expected frequency.
• In this problem I am applying Association rule with certain level of support and confidence threshold so that I can
generate rules which can help me to club product in such way that business can make use of it and sale more
products to customers
SUPPORT AND CONFIDENCE

• Threshold values of Support is 0.1


• Threshold values of Confidence is 0.4

Association Rules must satisfy both a minimum support threshold and a minimum confidence threshold.
KNIME WORKFLOW
ASSOCIATIONS

• Associations Identified --> Put the associations in a tabular manner --> Explain about support,
confidence, & lift values that are calculated
ASSOCIATION TABLE
SUPPORT ,CONFIDENCE AND LIFT CALCULATIONS
IN ASSOCIATION TABLE

• The number of transactions that include items in the {X} and {Y} then
• Support = (X+Y) / total
• Conf(X=>Y) = Supp(XU Y) / Supp(X)
• Lift(X=>Y) = Conf(X=>Y) / Supp(Y)
SUPPORT, CONFIDENCE, & LIFT VALUES  FOR
ASSOCIATION

• Rule 0 has support of .139 means approx. 13.9% Fraction of transactions containing the itemset hand soap , confidence of .401 means Probability of
occurrence of sandwich bags given hand soap is present is 40.1% and lift of 1.09, lift ratios higher than 1 indicate strong association between items.
• Rule 1 has support of .139 means approx. 13.9% Fraction of transactions containing the itemset hand soap , confidence of .401 means Probability of
occurrence of soda given hand soap is present is 40.1% and lift of 1.026, lift ratios higher than 1 indicate strong association between items.
• Rule 2 has support of .139 means approx. 13.9% Fraction of transactions containing the itemset hand soap , confidence of .401 means Probability of
occurrence of aluminum foil given hand soap is present is 40.1% and lift of 1.043, lift ratios higher than 1 indicate strong association between items.
• Rule 3 has support of .139 means approx. 13.9% Fraction of transactions containing the itemset hand soap, confidence of .401 means Probability of
occurrence of toilet paper given hand soap is present is 40.1% and lift of 1.06, lift ratios higher than 1 indicate strong association between items.
• Rule 4 has support of .139 means approx. 13.9% Fraction of transactions containing the itemset hand soap, confidence of .401 means Probability of
occurrence of cheeses given hand soap is present is 40.1% and lift of 1.026, lift ratios higher than 1 indicate strong association between items.
• Rule 5 has support of .139 means approx. 13.9% Fraction of transactions containing the itemset hand soap, confidence of .401 means Probability of
occurrence of individual meals given hand soap is present is 40.1% and lift of 1.026, lift ratios higher than 1 indicate strong association between items.
RECOMMENDATIONS

• Suggestion of Possible Combos with Lucrative Offers --> Write recommendations --> Make
discount offers or combos (or buy two get one free) based on the associations and your
experience
RECOMMENDATIONS

• Customer who are buying Hand soap can buy sandwich loaves too
• Customer who are buying sandwich loaves can buy Butter too
• Customer who are buying flour can buy all purpose too
• Customer who are buying flour can buy sandwich loaves too
• Store can run Combo offers like hand soap and ketchup , hand soap toilet paper, soda and sandwich loaves , pasta and
sandwich loaves these have high chance of customer buying together which will increase/boost the sales of product
• Store can also run offer life buy too and get one free or buy three and get one free like buy cheeses, soda and sandwich
bags and get hand soap free. There all items have strong lift in association rule means customer can go for buying
these item together and it will boost the product sales
Thank You

You might also like