HYBRID MACHINE LEARNING MODEL FOR FRAUD DETECTION IN UPI PAYMENT SYSTEMS

Author

G. Deshna , Dr. Nancy Jasmine Goldena

Abstract

 The growing use of Unified Payments Interface (UPI) has made digital payments easier and more convenient, but it has also increased the risk of fraudulent transactions. This paper studies the use of machine learning for detecting fraud in UPI transactions by comparing four supervised learning algorithms: Logistic Regression, Decision Tree, Random Forest, and XGBoost. Before training, the transaction data is cleaned and prepared using one-hot encoding, feature scaling, train-test splitting, and SMOTE for handling class imbalance. The four models are evaluated using Accuracy, Precision, Recall, F1-Score, and ROC-AUC to compare their fraud detection performance. The results show that XGBoost gives the best overall performance among the four models and is selected for the final prediction task. Additionally, an XGBoostRF hybrid model is included to extend the comparative analysis and investigate an alternative approach to UPI fraud detection. Its performance is compared with the original XGBoost model to understand the differences in their fraud detection capabilities. SHAP is then used with XGBoost to explain the features that influence individual transaction predictions. This study shows that combining machine learning with Explainable Artificial Intelligence can support fraud detection while also making the model's decisions easier to understand.

Keywords

UPI, Fraud Detection, Machine Learning, XGBoost, XGBoostRF, Random Forest, Decision Tree, Logistic Regression, Explainable AI, SHAP, Flask, Spring Boot.

Full Text:

Download Paper PDF

References

[1] T. Chen and C. Guestrin, “XGBoost: A Scalable Tree Boosting System,” in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 785–794, 2016.

[2] D. S. Nijwala, S. Maurya, M. P. Thapliyal, and R. Verma, “Extreme Gradient Boost Classifier Based Credit Card Fraud Detection Model,” in Proceedings of the IEEE International Conference on Device Intelligence, Computing and Communication Technologies (DICCT), pp. 500–504, 2023.

[3] J. West and M. Bhattacharya, “Intelligent Financial Fraud Detection: A Comprehensive Review,” Computers & Security, vol. 57, pp. 47–66, 2016.

[4] A. L. Buczak and E. Guven, “A Survey of Data Mining and Machine Learning Methods for Cyber Security Intrusion Detection,” IEEE Communications Surveys & Tutorials, vol. 18, no. 2, pp. 1153–1176, 2016.

[5] S. Misra, S. Thakur, M. Ghosh, and S. K. Saha, “An Autoencoder Based Model for Detecting Fraudulent Credit Card Transaction,” Procedia Computer Science, vol. 167, pp. 254–262, 2020.

[6] S. Dash, S. Das, S. Sivasubramanian, N. Kalyana Sundaram, G. K. Harsha, and T. Sathish, “Developing AI-based Fraud Detection Systems for Banking and Finance,” in Proceedings of the 5th International Conference on Inventive Research in Computing Applications (ICIRCA), pp. 891–897, 2023.

[7] B. Sethi, S. Mhatre, S. Yadav, S. Das, and V. Jadhav, “Machine Learning-Based UPI Fraud Detection: A Comprehensive Approach Using Random Forest,” in Proceedings of the MULTINOVA: First International Conference on Artificial Intelligence in Engineering, Healthcare and Sciences (ICAIEHS), pp. 462–470, 2025. DOI: 10.2991/978-94-6463-852-3_29.

[8] M. V. Sree, P. Bhuvaneshwari, K. Nandini, K. R. P. Amulya, P. Swetha, and B. Chandrakala, “Enhanced Detection of UPI Frauds Using Advanced Machine Learning Techniques for Secure Digital Transactions,” in Proceedings of the 4th International Conference on Information Technology, Civil Innovation, Science, and Management (ICITSM), 2025. DOI: 10.4108/eai.28-4-2025.2357830.

[9] N. Siva, K. Durgadevi, J. Naga Jyothi, D. Dadakhalandhar, K. Guruteja, and M. Keerthan Reddy, “A Machine Learning Framework for Detecting Fraudulent Transactions in UPI Systems,” in Proceedings of the 2025 conference, 2025.

[10] G. Ramya, A. Naga Venkata Gayathri, H. Althaf, and M. I. K. Shariff, “Explainable UPI Fraud Detection Using XGBoost and SHAP-Based Interpretable AI,” International Journal of Engineering Research and Science & Technology, vol. 22, no. 1, pp. 289–295, 2026. DOI: 10.62643/ijerst.2026.v22.n1(2).pp289-295.

Share your valuable work from Social Media Buttons