Title: k-Means clustering-based evolutionary algorithm for solving optimisation problems

Authors: Tribhuvan Singh; Krishn Kumar Mishra; Ranvijay

Addresses: Department of Computer Science and Engineering, ITER, Siksha 'O' Anusandhan (Deemed to be University), Bhubaneswar, India ' Department of Computer Science and Engineering, MNNIT Allahabad, Prayagraj, India ' Department of Computer Science and Engineering, MNNIT Allahabad, Prayagraj, India

Abstract: Environmental adaptation method (EAM) is a newly developed optimisation algorithm for complex problems. Although EAM and its variants converge very fast in lower-dimensional problems, the performance of these algorithms falls drastically in higher-dimensional problems. This paper introduces a novel approach to improve the performance of the algorithm in higher-dimensional problems. In order to explore the whole search space, the problem search space is divided into multiple mutually exclusive clusters, and then parallel exploitation and exploration are achieved that produces better results. The solutions of independent clusters try to adopt a more suitable structure using the direction received from the local/global best and local/global worst solutions. The performance of the suggested algorithm is compared with other existing algorithms using the benchmark function of the COmparing Continuous Optimisers (COCO) framework. The experimental results have demonstrated that the proposed algorithm performs well in many ways.

Keywords: evolutionary algorithms; optimisation problems; EAM; environmental adaptation method; k-Means clustering; parallel exploitation and exploration.

DOI: 10.1504/IJFE.2021.118911

International Journal of Forensic Engineering, 2021 Vol.5 No.2, pp.87 - 101

Received: 26 Aug 2020
Accepted: 12 Nov 2020
Published online: 11 Nov 2021 *

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