Fraudulent E-Commerce Transaction Detection Using Ensemble Machine Learning & Deep Learning

Online retail and e-commerce have grown exponentially over the past decade, offering convenience and accessibility to consumers worldwide. However, this rapid expansion has been accompanied by a surge in fraudulent transactions, posing significant challenges to both businesses and financial institutions. In response, this project proposes a robust fraud detection system that leverages advanced machine learning and deep learning techniques. By combining traditional classifiers—such as Logistic Regression, Random Forest, and XGBoost—with a deep neural network, and ultimately fusing their outputs through an ensemble approach, the system aims to improve detection accuracy and reduce false positives. Comprehensive data preprocessing, advanced feature engineering, and effective handling of class imbalance are central to this approach. Furthermore, the model has been deployed via a Streamlit-based interface, making real-time fraud detection accessible and interactive for end users. This report details the system architecture, methodologies, experimental evaluations, and future directions for research and development.