Applying machine learning in predicting backorder: An onpoint e-commerce company case study

Abstract

OnPoint is a company specialized in the field of E-commerce partnering. Their main business includes store commercial and operations, digital marketing, customer services, warehousing and fulfillment. Having to manage a high number of stock keeping units (SKUs) with various brands on different selling platforms proved to be quite a mountain to climb for a company which would celebrate only its 3rd birthday this December. Due to the huge number of SKUs and a relatively limited warehousing management system, backorders happen quite frequently. The purpose of this thesis is to develop a machine learning - based model which would act as a backorder risk detector, pushing information timely & correctly to managers in order to make the right decisions replenishment-wise. Ensemble machine learning methods were the main techniques used to solve this problem. Keywords: backorder prediction; machine learning; E-commerce; ensemble learning; warehouse management