Holistic swarm optimization: A novel metaphor-less algorithm guided by whole population information for addressing exploration-exploitation dilemma

This paper introduces a novel metaphor-less optimization algorithm called Holistic Swarm Optimization (HSO), designed to enhance the search process by utilizing data from the entire population. Unlike conventional algorithms that rely on partial or local information, HSO adopts a comprehensive approach, ensuring that each decision is informed by the overall distribution and fitness landscape of the population. The algorithm dynamically balances exploration and exploitation through an adaptive framework that integrates root-mean-squared (RMS) fitness-based displacement coefficients, simulated annealing-based selection, and adaptive mutation. This structure enables HSO to efficiently navigate complex, multimodal optimization problems while avoiding local optima. The performance of HSO is evaluated on two widely used benchmark test suites–CEC 2005 and CEC 2014–and a series of real-world engineering design problems. Results show that HSO delivers competitive and stable performance when compared to several state-of-the-art metaphor-based and metaphor-less algorithms. These findings demonstrate the effectiveness of a holistic population-guided approach in achieving robust optimization outcomes, making HSO a promising alternative for solving diverse and challenging problems without reliance on metaphorical inspirations. The source codes and implementation guidance for the HSO algorithm are available for public access on the https://github.com/ebrahimakbary/HSO.