Customer Behavior Prediction using Deep Learning Techniques for Online Purchasing

E-commerce success may be improved with a deeper knowledge of how customers make purchases online. The primary emphasis of prior research has been on consumers’ intent to buy, with sales rank serving as a proxy. Furthermore, there is no guarantee that a shopper’s retail buy intent would translate into action. Our goal is to get a deeper comprehension of how online shoppers use an e-commerce platform by anticipating their actions based on a huge multidimensional data sample consisting of more than 50,000 individual web sessions. In this investigation, we identified two groups of variables—platform engagement and customer characteristics—as significant predictors of retail consumers’ propensity to shop online. We also compared our deep learning approach’s predictive efficacy to that of some of the most popular supervised learning methods for forecasting, such as Decision Trees, Support Vector Machines,Random Forests, and ANNs. When compared to other machine learning methods, the deep learning approach performed better on the same dataset. These studies will aid in the planning stages of platform creation, and they will also contribute to the growth of the academic literature on the topic of purchase forecasting for online e-commerce platforms.