Basket Analysis helps us to understand the buying relationship between two product categories. In other words, basket analysis helps us to understand which two products customers are more likely to purchase together.
I have done basket analysis for Product Categories. It helps us understand which two product categories are consumers more likely to purchase together. It is important to note that in the matrix, there is mismatch between "A & B" compared to "B & A" combination. This is true to design.
34.2% of Customers buying from “Condiments” category will most likely purchase from "Confections" category. On the other hand, only 22.37% of customers buying “Confections” likely to buy “Condiments”.
In this project, there are few measures calculated:
Orders Both - Number of orders where combination are products match
Order Support - % of Orders containing a particular combination compared to total orders
Confidence Level - How much likely customer who purchase one kind of product/product category will more likely purchase product from another category
Lift - It refers to ratio of confidence and fraction of products for a combination. It helps to identify whether the relationship is strong enough to make a prediction.