MRA Project 2: Sudesh Yadav
MRA Project 2: Sudesh Yadav
Sudesh Yadav
• Problem statement
Agenda: • Data description
• Tools Used
• Exploratory Analysis
• Market Basket Analysis
• Rules identified
• Recomendations
Problem A Grocery Store shared the transactional data with you. Your
job is to identify the most popular combos that can be
statement: suggested to the Grocery Store chain after a thorough analysis
of the most commonly occurring sets of items in the customer
orders. The Store doesn’t have any combo offers.
•Hand Soap is a product that ordered least among the product list. And is part of 394 orders out of 1139 orders.
Followed by Sandwich loaves and flour and so on as below:
EDA - Orders trend over the years
•The orders trend for the
given data is decreasing over
the years with 2018 having
highest orders and then
followed closely by 2019 and
then 2020.
•2019
•2020
•Category 1 – NC (Non-
cosumables like soap and liquid
detergent etc.
• In our case as an example the product paper towels being bought along with set of
[eggs, dinner rolls, ice cream, pasta, lunch meat] is
2.349 times higher when compared to it being bought individually which is just 0.02.
• Like this with a minimum of below threshold values the association rules for the given data are calculated:
Support of minimum = 0.02 (This is the minimum possibility of a product being ordered)
Maximum Item set length = 10 ( Maximum counts of products in 1 set of the order basket)
Minimum Confidence level = 0.08 (Minimum confidence that the product suggested would be
picked up while the other set of products are in the basket)
MBA – Association Rules in tabular form
39 such Association rules or suggestions :
MBA – KNIME workflow used for MBA
MBA – Association Rule Parameters
The threshold values are found out by
various regressions and shown here:
Ssuggestions & Recommendations
• Poultry could be suggested as a combo offer with most of the food and snacks items such as dinner rolls and
spaghetti sauce.