A hybrid binary Grey Wolf and Whale optimization method for feature selection in classification tasks

Siswantoro, Joko and Wijaya, Jabesh Nehemiah and Prasetyo, Vincentius Riandaru and Arwoko, Heru (2026) A hybrid binary Grey Wolf and Whale optimization method for feature selection in classification tasks. Expert Systems with Applications, 231. p. 132277. ISSN 1873-6793 (In Press)

	PDF 1-s2.0-S0957417426011905-main published.pdf - Accepted Version Restricted to Registered users only until January 2027. Download (4MB) | Request a copy
	PDF Expert Systems with Applications halaman judul editorial daftar isi.pdf - Other Download (205MB)
	PDF Bukti corresponding author eswa.pdf - Other Download (1MB)

Abstract

In recent years, the growing volume of high-dimensional data has posed significant challenges to machine learning tasks, particularly feature selection, where removing irrelevant or redundant features is critical to improve model performance and reduce computational cost. This study proposes a new Binary Grey Wolf-Whale Optimization (BGWWO), a hybrid metaheuristic algorithm that merges the exploration ability of the Grey Wolf Optimizer (GWO) and the exploitation ability of the Whale Optimization Algorithm (WOA), optimized by a novel reverse-V-shaped transfer function for binary optimization. BGWWO is experimented on feature selection tasks for ten benchmark datasets, with K-Nearest Neighbors (KNN) as the classifier. Careful experiments were conducted to compare the performance of the proposed method with established algorithms, i.e., binary versions of Particle Swarm Optimization (PSO), GWO, Salp Swarm Algorithm (SSA), and WOA. The results show that BGWWO outperforms competing algorithms significantly in terms of classification accuracy, number of selected features, fitness value, computational time, and memory usage, providing a better balance between predictive capability and dimension reduction. These findings show the potential of BGWWO as a robust and effective feature selection technique for a broad variety of classification problems, providing a valuable contribution to the field of metaheuristic optimization research.

Item Type:	Article
Uncontrolled Keywords:	Binary Grey Wolf-Whale optimizationFeature selectionMetaheuristic algorithmsClassification accuracyDimensionality reduction
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:	Faculty of Engineering > Department of Informatic
Depositing User:	Joko Siswantoro
Date Deposited:	14 Apr 2026 02:04
Last Modified:	14 Apr 2026 02:04
URI:	http://repository.ubaya.ac.id/id/eprint/50535

Actions (login required)

View Item