Machine learning pipeline for online shopper intention classification

Abstract

Nowadays people prefer to buy online rather than buy on the spot. Online transactions make people's lives easier. Tins condition requires the seller to understand the characteristics of the intention of the prospective buyer. This research proposes a machine-learning pipeline to predict the customer behavior for e-commerce products. We compared several machine-learning algorithms to find the best algorithm to solve the problem then we deployed the model on a web application. Based on the experiment, the Random Forest algorithm can predict online shopper intention with 90% of accuracy.