Leveraging Machine Learning Techniques of Real Time Detection of UPI Fraud

Quickly changed the status of digital transactions to record the Unified Payments Interface (UPI). At the same time, it also increased users' sensitivity to financial fraud. In this document that balances the same, this study recommends an effective fraud detection system using a machine learning approach. The proposed system effectively identifies rogue transactions by pursuing transaction behavior and user activity. Experimental results show that the accuracy of the model based on logistics regression is 97%. Comparative analysis with decision trees, random forests, and Naive Bayesian models checks for excellent performance of logistics regression related to accuracy and recirculation measurements. The system also provides interpretation results by marking illicit transactions and creating timely warnings. Flexible solutions improve security of UPI and guarantee digital payment systems.