Accurate Prey Localization (APL) for Promoting the Performance of Chimp. Bio-Inspired Optimization Techniques

Document Type : Original Article

Authors

1 Computer and Systems Engineering Department, Faculty of Engineering, Horus University, New Damietta, Egypt

2 Computers Science Department, Arab East Colleges, Saudi Arabia

3 Computers Engineering and Control Systems Department, Faculty of Engineering, Mansoura University, Mansoura, Egypt

Abstract

Bio-inspired optimization algorithms have emerged as powerful computational tools for solving complex optimization problems, with the Chimp Optimization Algorithm (ChOA) representing a notable advancement through sophisticated cooperative hunting behaviors. However, a critical limitation exists in ChOA's prey localization mechanism, which treats all leading solutions equally when estimating prey position regardless of their individual fitness levels, leading to inaccurate positioning and suboptimal conver-gence. This research introduces Accurate Prey Localization (APL), a novel enhancement that replaces traditional simple averaging with fitness-aware weighted positioning, im-plementing a sophisticated pairwise estimation strategy where chimps with higher fit-ness values receive proportionally greater influence in prey position calculations. Exper-imental validation demonstrates substantial effectiveness of APL-Improved ChOA compared to original ChOA and Grey Wolf Optimization across multiple evaluation metrics, achieving approximately 15,000-fold improvement in convergence perfor-mance. A comprehensive case study on COVID-19 feature selection further validates APL's practical effectiveness, with Binary APL-Improved ChOA achieving 98% accuracy compared to competing algorithms' maximum of 94.1%. The Improved convergence be-havior, improved solution accuracy, and consistent performance establish APL as a sig-nificant advancement for bio-inspired optimization algorithms, particularly valuable for complex engineering design problems, feature selection tasks, and high-dimensional optimization scenarios where traditional methods struggle with local optima entrap-ment.

Keywords

Files

Share

How to cite

Statistics