Whale Optimisation Algorithm for high-dimensional small-instance feature selection. Issue 2 (4th March 2021)

• Record Type: Journal Article
• Authors: Mafarja, Majdi; Jaber, Iyad; Ahmed, Sobhi; Thaher, Thaer
• Is Part Of: International journal of parallel, emergent and distributed systems. Volume 36:Issue 2(2021)
• Page Start: 80
• Page End: 96
• Publication Date: 2021-03-04
• Subjects: Whale Optimisation Algorithm -- WOA -- transfer functions -- high-dimensional data -- classification -- feature selection; Parallel computers -- Periodicals; Electronic data processing -- Distributed processing -- Periodicals; Computer algorithms -- Periodicals; 004.35
• DOI: 10.1080/17445760.2019.1617866
• Languages: English
• ISSNs: 1744-5760

Abstract

Abstract : In this paper, eight variants of the Whale Optimisation Algorithm (WOA), that are based on eight different transfer functions, are introduced and used as search strategies in a wrapper feature selection model. Feature selection is a challenging task in machine learning process. It aims to minimise the size of a dataset by removing redundant and/or irrelevant features, with no information loss, to improve the efficiency of the learning algorithms. The used transfer functions belong to two different families; S-shaped and V-shaped. The proposed approaches have been tested on nine different high-dimensional medical datasets, with a low number of samples and multiple classes. The results revealed a superior performance for the V-shaped based approaches over the the S-shaped approaches. Moreover, the results of the V-shaped approach is compared with well-known feature selection approaches, and the superiority of the proposed approach is proven. GRAPHICAL ABSTRACT: