UPI Fraud Detection

The widespread adoption of the Unified Payments Interface (UPI) has introduced a new level of convenience in digital transactions. However, this growth has been accompanied by an alarming rise in fraudulent activities, such as phishing, unauthorized transactions, and identity theft. This project presents an advanced UPI fraud detection system to address these challenges. Using HGBoost—a hybrid gradient boosting model—and deep learning techniques, the system analyzes transaction data, user behavior, and anomalies in real-time. The proposed model combines decision trees with boosting algorithms to improve detection accuracy while reducing false positives. The system's capability was validated on a dataset of UPI transactions, demonstrating high precision and recall rates. This ensures transaction security and preserves the speed and seamless experience expected from UPI systems.