A Predictive System for Fraud Detection in Ecommerce Transactions Using Classification Models

Alva-Cordova F.; Reyes-Suarez L.; Bautista-Ubillús E. · 2025 IEEE International Conference on Quantum Photonics, Artificial Intelligence, and Networking, QPAIN 2025 (2025) · DOI: 10.1109/qpain66474.2025.11172218

Resumen

This paper proposes a predictive system for fraud prevention in e-commerce transactions, leveraging machine learning classification and facial recognition technologies. We created a specialized dataset tailored to fraudulent and legitimate transaction patterns and evaluated several classification models for optimal performance in fraud detection. Our experiments show that the Random Forest classifier outperforms other models in accuracy and precision. By integrating this highly effective model with facial recognition technology, the proposed system enhances security and mitigates fraud risk in e-commerce platforms, achieving 98% precision and recall using Random Forest, which significantly enhances user trust in online transactions.

Instituciones: Universidad Peruana de Ciencias Aplicadas