Predicting Online Purchases Using Six Machine Learning Models Based on Customer Demographics

This research paper investigates the predictive power of six machine learning models based on customer demographics to classify online purchasing behavior. The data set consisting of 1000 customer transactions was compared against six various classification algorithms: Naïve Bayes, Decision Tree, Random Forest, K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Logistic Regression. Comparative evaluation revealed that the Random Forest model outperformed all the rest of the classifiers and achieved a remarkable accuracy of 91.5 %, thereby asserting its superiority and efficacy regarding predicting buying behavior. The study highlights the significance of demographic factors, including age, income, education, and gender, as determinants of online buying behavior. Furthermore, it delves into actual applications, ethical issues, and avenues of future research, all showing the vast scope of predictive analytics in modeling consumer behavior. With predictive insights facilitated through machine learning, companies can optimize their marketing campaigns, customer targeting, and overall user experience. This research contributes to the field of e-commerce analytics and provides a foundation for further research to examine behavioral and psychographic indicators to enhance predictions. This paper presents original research through comparative analysis of machine learning classifiers and does not serve as a review article.