Optimization-Based Machine Learning Models for Credit Card Fraud Detection: A Systematic Review with Emphasis on Cuckoo Optimization Algorithm

Detection of credit card fraud has become a major concern in the contemporary financial structures with the fast development of online payment systems and electronic transactions. Conventional methods of fraud detection are not usually very accurate in case of the large scale datasets that are highly unbalanced. Optimization algorithms have recently gained popularity to be used together with machine learning models to enhance detection performance, feature selection, and classification accuracy. The paper has carried out a methodical review of optimization-based methods in credit card fraud detection systems with particular consideration of Cuckoo Optimization Algorithm (COA). The general search of the literature was carried out in the key scientific databases such as Scopus, IEEE Xplore, ScienceDirect, etc.