I'm working with data where many customers only buy 1-3 products at a time, meaning that there aren't enough products being purchased together for the market basket algorithm to determine associations. Any idea how I can get around this? I'm thinking of grouping transactions together by week or month to get larger basket sizes, but I'm skeptical of that approach since customers can place many orders in a week that have nothing to do with each other.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.