UPI FRAUD DETECTION SYSTEM

Abstract

Now a days Digital transactions are rapidlyincreasing as it results in increasing online paymentfrauds too. In fact, according to the Reserve Bank ofIndia, comparing March 2022 to March 2019,digital payments have risen in volume and value by216% and 10%, respectively. People are starting togo all-in with digital transactions, but one can’tdeny the security issues that loom, and know-howwhen it comes to online payments. Few years ago,we could have barely seen the online payment, buttoday UPI payment QR code installed at doorstep.This invited the hoaxers and attackers to developfraudulent transactions and fool people for someamount of money. Fortunately, the onlinetransactions are monitored and hence could beanalysed using the latest tools. In this system, anattempt is made to develop a machine learningmodel to identify fraudulent transactions in atransaction’s dataset. This study considers the taskof applying artificial intelligence to recognize bankfraud. In recent years, due to the COVID-19pandemic, bank fraud has become even morecommon due to the massive transition of manyoperations to online platforms and the creation ofmany charitable funds that criminals can use todeceive users. The study’s scientific novelty is thedevelopment of machine learning models foridentifying fraudulent banking transactions andtechniques for preprocessing bank data for furthercomparison and selection of the best results. Thispaper also details various methods for improvingdetection accuracy, i.e., handling highly imbalanceddatasets, feature transformation, and featureengineering. The recognition of banking fraud usingartificial intelligence algorithms is a topical issue inour digital society.