Advances in Feature Selection Using Memetic Algorithms: A Comprehensive Review

Abstract

This review paper presents a comprehensive analysis of the memetic algorithms (MAs) for feature selection (FS), particularly in high‐dimensional datasets. MAs effectively address the challenges of feature selection by combining the global exploration capabilities of evolutionary algorithms with the local optimization of search techniques. Their hybrid nature makes them well suited for tackling the complexity, scalability, and computational demands of FS problems across various domains, including bioinformatics, image processing, and financial forecasting. This review highlights the recent advancements, customized variants, and practical applications of MA‐based FS methods while providing critical insights into their limitations, such as computational overhead and overfitting. Additionally, the paper outlines future research directions to further enhance the efficacy of MAs in feature selection, offering a balanced perspective on their contributions to the field.