Title: Meta-heuristics for feature selection: a comprehensive survey and comparative analysis

Authors: Rishika Kumar; Ashish Jain; Inderjeet Kaur

Addresses: Department of Information Technology, School of Information Technology, Manipal University Jaipur, Jaipur, 303007, India ' Department of Information Technology, School of Information Technology, Manipal University Jaipur, Jaipur, 303007, India ' Department of Computer Science and Engineering, Galgotias College of Engineering and Technology, Greater Noida, 201310, India

Abstract: Feature selection (FS) is a crucial step in pre-processing of data that aims to identify a subset of relevant features from a large pool of available features, while discarding irrelevant or redundant ones. From early 2000s, optimisation heuristic methods have gained popularity as an alternative to traditional FS methods. In the literature, it has been shown that the optimisation heuristics can efficiently search for a subset of relevant features that can represent the data accurately. They are flexible, scalable, and can handle non-differentiable objective functions, making them suitable for FS. In this paper, we comprehensively review those optimisation heuristics that have been developed in last one decade and applied successfully for FS. Each algorithm is elucidated theoretically, providing in-depth explanations of their methodologies. This survey presents difficulties faced by optimisation heuristic FS algorithms and prospective research directions are analysed and highlighted for the benefit of researchers working in this area.

Keywords: feature selection; optimisation heuristics; data accuracy.

DOI: 10.1504/IJBRA.2024.141378

International Journal of Bioinformatics Research and Applications, 2024 Vol.20 No.4, pp.323 - 356

Received: 15 Nov 2023
Accepted: 22 Jan 2024
Published online: 10 Sep 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article