Machine Learning-Based UPI Fraud Detection

Abstract

Despite the eventuality for a number of cheat conditioning, the Unified Payments Interface (UPI) has altered the online sale geography in India.This composition provides a quick review of machine learning( ML) ways for UPI fake discovery. Machine learning( ML) techniques dissect sale data using a combination of supervised, unsupervised, and semi-supervised learning ways to find anomalous patterns that may indicate fraudulent gesture. Effective point selection and engineering strategies are pivotal for maximizing the performance of the models. The use of original-time monitoring systems and adaptive literacy ways increases the adaptability of UPI security and enables briskly responses to arising fraud ways. By using machine learning capacities, fiscal institutions can maintain the integrity of UPI deals and ameliorate their security while erecting client confidence.

Introduction

To tackle this issue, the study proposes a machine learning-based fraud detection system, focusing on the Random Forest algorithm. Random Forest, an ensemble learning method, combines multiple decision trees to analyze transaction data and detect complex patterns or anomalies that may indicate fraudulent activity. It is chosen for its high accuracy and robustness in handling large, complex datasets.

The system is built using a dataset sourced from Kaggle, with data preprocessing involving an 80/20 train-test split. The model is trained using supervised learning to classify transactions as legitimate or fraudulent.

The results highlight that machine learning significantly improves fraud detection by:

• Identifying suspicious patterns in real time
• Continuously learning and adapting to new fraud techniques
• Enhancing accuracy and scalability in high-volume transaction environments

Overall, the proposed system strengthens UPI transaction security by providing efficient, adaptive, and real-time fraud detection, helping protect users and the financial ecosystem from evolving cyber threats.

Conclusion

Random Forests and sorted due to its versatility, which enables it to effectively manage asymmetric datasets and a variety of data sources. farudsters’confidence in UPI security is increased as a result of the system’s Rapid fire discovery of abnormalities that signify fraudulent geste, which greatly increases the responsibility of digital deals.

References

[1] Narendra Kumar, etal.(13-10-2020). productoverviewhttps//www.npci.org.in/what-we-do/upi/product-overview [2] Chatterjee D.A. and Thomas R.( 2017). Unified payment interface( upi)A catalyst tool supporting digitalization–mileage,prospects and issues,International Journal of Innovative exploration and Advanced Studies( IJIRAS), vol. 4, no. 2, pp. 192 – 195. [3] Lakshmi K., Gupta H., and RanjanJ.(2019).Upi grounded mobile banking operations–security analysis and advancements, in 2019Amity International Conference on Artificial Intelligence(AICAI), pp. 1 – 6. [4] Eldefrawy M.H., Alghathbar K., and KhanM.K.(2011).Otp-grounded two factor authentication using mobile phones, in 2011 Eighth International Conference on Information Technology New Generations,pp. 327 – 331. [5] Base-Bursey M.(26-11-2020).What are app warrants-a look into android app permissions.https//www.wandera.com/mobile-security/app-and-data-leaks/app-permissions/. [6] Dr. Virshree Tungare.(2019).A study on client sapience towards upi(unifiedpaymentinterface)-an advancement of mobile payment system, International Journal of Science and Research( IJSR). [7] Mohapatra S., et al. (2017). Unified payment interface (upi):a cashless indian transaction process., International Journal of Applied Science and Engineering, vol. 5, pp. 29–42, 6.

Copyright

Copyright © 2026 Farooque Alam, Nikhil Rajput. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.