Review of Advances in Fuzzy Logic Models for Fraud Detection

Abstract

Abstract

The digital payment landscape has transformed considerably over the years, with more people choosing electronic payment platforms over traditional banking methodologies. Numerousvirtual banking services trends have emerged to make transactions seamless for customers. However, the increase in electronic fraud presents a significant threat to Nigeria's financial system, causing substantial economic losses and impeding the broader adoption of digital transaction channels. This study attempt to review the advances made in using fuzzy logic for fraud detection as well as their challenges. The findings indicate fuzzy logic dominance in fraud detection. However, there is a noticeable concentration on narrow range of transactions types and use of imbalance data. The study stressed the need for hybrid fuzzy logic system that integrates data from various transaction types, and include machine learning to handle emerging transaction types such as cryptocurrency. The paper also proposed a fuzzy logic prototype for effective monitoring and fraud prevention across various channels.