Business Problem
The goal of this project is to predict the missing items in a shopping cart and apply
association rules. Based on the frequently occurring transactions and the items currently in
a shopping cart, other items that a customer is likely to add before proceeding to the
checkout counter can be predicted. For example, if a customer has two food items in the
shopping cart which are toothbrush and toothpaste, it is possible to predict that the
customer can add mouthwash to the shopping cart before checkout. In our work, we
wanted to make the next logical step by allowing any item to be treated as a class label and
its value is to be predicted based on the presence or absence of other items. Put another
way, knowing a subset of the shopping carts contents, we want to predict the rest. Suppose
the shopping cart of a customer at the checkout counter contains toothbrush, toothpaste,
mouthwash, could someone who met the same customer when the cart contained only
toothbrush, toothpaste, mouthwash, have predicted that the person would add body wash
and shampoo? Anticipating reorder levels for products is the second objective. Based on
the products that are present in a customer’s shopping cart, we can predict future sales. In
predicting future sales, the store must make sure that there is sufficient inventory level to
meet the predicted future sales level. When a product reaches a minimum level, it must be
reordered to keep inventory levels consistent with the minimum reorder level.
Introduction – Association Rules
Association Rules represent an unsupervised learning method that attempts to capture
associations between groups of items. Association Rules have also been referred to in the
literature as Market Basket analysis or Affinity analysis. . Market Basket analysis aims at
discovering which groups of products tend to be purchased together. There is a two stage
process involved- rule generation and assessment of rule strength to choose a subset.
Association rules or affinity analysis, constitute a study of “what goes with what”. More
formally an item list is a set of items chosen during a given trip to a store while a
transactions file is a data file that details the items purchased by customers during trips to
the store. Association rules are used to find out the interesting and useful relationships
between data that occur frequently enough to be called a pattern (or a trend) and hence, can
be formulated into a rule. Each of these rules has an if-then structure with an antecedent
