Binary glider snake optimization for feature selection

Seyyid Ahmed Medjahed

doi:10.1007/s10586-026-06427-9

Return

Binary glider snake optimization for feature selection

2026-08-17

PRE

Seyyid Ahmed Medjahed *

Rahma Lama

DOI：10.1007/s10586-026-06427-9

Original

Original request for help

Save

Abstract

En 中文

Feature selection plays a crucial role in machine learning applications involving high-dimensional datasets, where redundant and irrelevant features may degrade predictive performance and increase computational complexity. This paper presents a binary adaptation of the Glider Snake Optimization (GSO) algorithm, termed Binary Glider Snake Optimization (BGSO), for feature selection problems. Inspired by the gliding and locomotion behavior of arboreal snakes, the proposed approach extends the continuous search mechanism of GSO to binary optimization. To improve the search process, BGSO incorporates a dual-guidance learning strategy that exploits information from both the best-performing solution and neighboring agents, facilitating effective information propagation within the population. In addition, an adaptive weak-agent replacement mechanism is employed to preserve population diversity and mitigate premature convergence. The proposed method is evaluated on several benchmark classification datasets and compared with representative binary metaheuristic algorithms. Experimental results, supported by statistical analyses, indicate that BGSO is capable of identifying compact feature subsets while maintaining competitive classification performance. These findings suggest that the proposed binary adaptation provides an effective alternative for addressing high-dimensional feature-selection problems.

Keywords:

Feature selection

Binary optimization

Metaheuristic algorithms

Glider snake optimization

Wrapper model