Fraud Detection with Ensemble Learning and Web-Based Implementation
Nahida Fatme, Md. Moynul Islam, Jahid Farhad Apon, Md Mahmudur Rahman · 2024
In the ceaseless battle against digital fraud, our fraud detection project emerges as a proactive and comprehensive response to the audacity of fraudsters who pilfered millions, showcasing the urgency to fortify defenses. Navigating the intricacies of big data, we harness the power of advanced Machine Learning, deploying a squad of eight classification models to combat the ever-evolving tactics of fraud. With an unwavering commitment to understanding the complexities of fraudulent activities, our objective is crystal clear: to contribute a high-performance fraud detection model. Facing the challenge of imbalanced data, we embark on a journey of exploration, identifying, and implementing resampling methods that significantly enhance the model’s efficiency. The crescendo of our efforts culminates in the deployment of a robust ensemble model, led by the exemplary Voting Ensemble, now accessible through a user-friendly website. This project not only achieves its core objectives but establishes a sustainable and efficient solution for real-world fraud detection, deftly bridging the gap between cutting-edge research and practical, impacting application.