COMBINING THE STRENGTHS OF RULE-BASED AND ANOMALY DETECTION TECHNIQUES FOR ROBUST AND COMPREHENSIVE PAYMENT FRAUD DETECTION

Abstract

Payment fraud is a pervasive problem that poses significant challenges for businesses and financial institutions worldwide. Traditional fraud detection methods often rely on rule-based systems or anomaly detection techniques, each with its own strengths and limitations. This research article explores the potential of combining the strengths of rule-based and anomaly detection techniques to create a robust and comprehensive payment fraud detection framework. By leveraging the domain knowledge and expert-defined rules of rule-based systems and the ability to identify novel fraud patterns through anomaly detection, the proposed hybrid approach aims to enhance the accuracy, adaptability, and scalability of fraud detection models. The article presents a detailed methodology, experimental results, and discusses the implications and future directions for payment fraud detection in the rapidly evolving digital landscape