E-Payment Fraud Detection in E-Commerce using Supervised Learning Algorithms
Manal Loukili, Fayçal Messaoudi, Hanane Azirar · 2024
In today’s e-business world, using credit cards has reinforced the development of e-commerce and facilitated the electronic payment system. Consequently, this excessive frequency of credit card use has favored increased fraud. Websites and e-commerce platforms operating with customers and handling sensitive user data must implement effective fraud prevention systems to detect and prevent fraud in e-payment transactions. Indeed, machine learning has proven effective in detecting and preventing fraud. In this context, this chapter aims to implement a machine learning system for detecting fraudulent e-payments. To do this, three supervised machine learning models have been compared, namely: CatBoost, AdaBoost, and XGBoost, based on their performance parameters (including precision, accuracy, recall, and F-score) and latency time, to build a powerful system that aims at detecting online fraud and avoiding losses caused by fraudulent transactions.