Title: Analysis of S-shaped, U-shaped and V-shaped transfer functions in IoMT datasets using binary Aquila optimisation techniques

Authors: Surendra Babu Nallagorla; R. Dhanalakshmi

Addresses: Department of CSE, Indian Institute of Information Technology Tiruchirappalli, Tiruchirappalli, 620012, India ' Department of CSE, Indian Institute of Information Technology Tiruchirappalli, Tiruchirappalli, 620012, India

Abstract: The Aquila optimisation (AO), a metaheuristic approach, is motivated by the Aquila's natural behaviour when collecting prey. However, while the method shines at several benchmark functions, it fails to solve the binary optimisation problem. We suggested a binary version of AO (BAO) for feature selection (FS) concerns in classification tasks in this research using internet of medical things (IoMT) datasets. We used 12 (S, U, and V-shaped) transfer functions (TF) to transform continuous data into binary values. The proposed TFs demonstrate that BAO techniques, particularly S2-BAO, outperform alternative transfer functions. According to the results, the suggested approach, as compared to traditional transfer functions, converges to the global minimum in multiple rounds based on the selection of optimal attributes, fitness values, and improved classification accuracy.

Keywords: Aquila optimisation; classification; feature selection; metaheuristic optimisation.

DOI: 10.1504/IJMOR.2023.134837

International Journal of Mathematics in Operational Research, 2023 Vol.26 No.3, pp.393 - 409

Received: 27 Jun 2022
Accepted: 17 Sep 2022
Published online: 14 Nov 2023 *

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