Volume 2 - Issue 2, March - April 2026

Online payment fraud detection is crucial to safeguarding e-commerce transactions from skilled criminals taking advantage of system vulnerabilities. This framework uses six machine learning algorithms to predict online payment fraud on three datasets: constant, CN7Rule induction, KNN, Tree, Random Forest, Gradient boosting, SVM, Logistic regression, Naive Bayes, Ada boost, neural network, and stochastic gradient descent. My extensive testing has shown that the particular gradient-boosting approach consistently outperforms other algorithms I have investigated. This algorithm's astounding accuracy of 99.7% is astounding. The algorithm became the best-performing framework for online payment fraud detection because of its resilience in a range of testing scenarios. The idea behind the invention is a gradient boosting first-aid solution for electronic fraud.

Mr.I.Gobi, Dhamu D,"Machine Learning Approaches for Online Payment Fraud Detection: Challenges and Performance Analysis" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(2): Page(439-443) Mar-Apr 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.