AI-Driven UPI Fraud Detection for Customer Safety

Banu, A.Aneesa; M.Geetha,; S.Dhivya,

doi:https://doi.org/10.55041/ijsmt.v2i4.202

Plagiarism Passed

Peer reviewed

Open Access

AI-DRIVEN UPI FRAUD DETECTION FOR CUSTOMER SAFETY

AUTHORS:

A.Aneesa Banu

M.Geetha

S.Dhivya

Mentor

D. Kalpana Devi

Affiliation

Dept. of CSE M.I.E.T Engineering College (Affiliated to Anna University) Trichy, TamilNadu

DOI: 10.55041/ijsmt.v2i4.202

CC BY 4.0 License:

This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

DOWNLOAD ARTICLE

REVIEW REPORT

CITE THIS ARTICLE

Abstract

— In this project, an artificial intelligence-based technique to detect fraud in Unified Payments Interface (UPI) networks to ensure that users’ transactions are protected is described. With the increased popularity of online payments, there is the need to detect such frauds in real-time to minimize any loss and increase the safety of consumers. Data used in this research are obtained and pre-processed to make sure the data does not contain any noise or null values, thus ensuring data accuracy. Various features associated with such transactions are identified and utilized to classify them. Using the identified features, a classifier using random forest algorithm is built and used to distinguish between fraudulent and non-fraudulent transactions. The random forest model enhances the performance of the classifier by using more than one decision tree hence improving prediction accuracy and avoiding overfitting. The proposed method can detect frauds in real- time hence helping to reduce the associated costs.

Keywords

Article Information

Article Metrics

Article Views

PDF Downloads

HOW TO CITE

References

Ethics and Compliance

✓ All ethical standards met

This article has undergone plagiarism screening and double-blind peer review. Editorial policies have been followed. Authors retain copyright under CC BY-NC 4.0 license. The research complies with ethical standards and institutional guidelines.

Indexed In