E-Commerce Fraud Detection Using Machine Learning

The rapid expansion of e-commerce has simultaneously increased the sophistication and frequency of fraudulent activities. This paper presents a machine learning-based approach for detecting e-commerce fraud using ensemble models-Stacking Classifier and XGBoost. A synthetic dataset of 23,634 transactions was generated to simulate realistic user and fraud behavior. The proposed system leverages Python and the Flask framework, offering real-time detection capabilities. Evaluation metrics such as precision, recall, and F1-score were used due to the highly imbalanced nature of fraud data. The results indicate significant improvement in fraud detection performance compared to traditional models, particularly in reducing false positives and negatives. The system shows strong adaptability to evolving fraud patterns and can be effectively integrated into e-commerce platforms for live transaction monitoring.