Utilizing Data Analytics to Analyze Online Purchase Behavior

Abstract

The emergence of data analytics has fundamentally transformed supply chain management strategies in the global marketplace during the past decade. Classification is one of the most popular methods and receives a great deal of attention in the literature, but there are still some questions concerning the performance characteristics of different classification methods. This paper analyzes three different classification methods: classification trees, k-nearest neighbors, and artificial neural networks to determine if there are any performance gaps between the methods. A series of experiments are conducted utilizing the Analytic Solver Data Mining (formerly XLMiner) add-in to Microsoft Excel in an effort to address these issues. The analysis reveals that there may be minor performance gaps, but the methods all perform well in the context of this study.