Office Depot: Sales Anomaly Detection

Sponsored by: Office Depot

 

Abstract
The goal of our project is to create subsystems identifying anomalies from three years of Office Depot online sales data. One of the program’s sub-systems determines if someone has placed too large or small of an order while the other program’s sub-system determines if the time between when a single user places two orders is too small. These programs will be implemented on a live data stream so that Office Depot employees can more easily rectify anomalies. Community Benefit: Our program will help Office Depot better handle its online sales by providing a second wall to better ensure improper sales don’t get pushed through for processing, which in turn will save the company money.


Team Members

Morgan Brady
Willie Miller
Dominick Mineo