Enhancing Fraud Detection in UPI Transactions using Ensemble Learning and Neural Networks

Abstract

The prevalence of Unified Payments Interface (UPI) transactions has become ubiquitous among individuals, who have come to rely on this method for numerous essential daily transactions. The utilization of UPI transactions is now widespread, encompassing small vendors to large firms. Individuals find UPI more convenient than alternative payment methods due to its ease of use. However, concurrent with the evolution of digitalization in online payments through UPI transactions, the risk of experiencing fraudulent activities has also increased proportionally. Perpetrators of fraud have enhanced their methodologies to manipulate and defraud unsuspecting individuals, resulting in financial losses. Traditional fraud prevention approaches, such as awareness programs, have proven insufficiently effective, as individuals continue to fall victim to the schemes devised by fraudsters, ultimately incurring monetary losses. Consequently, there is a significant need for solutions that can contribute to loss prevention before fraudsters succeed in causing disruptions and adversely affecting individuals' financial status.