Human-inspired metaheuristic algorithms: a comprehensive review of theory, design, and applications

Abstract

Metaheuristic algorithms are indispensable for solving complex optimization problems that challenge traditional methods. One of the subclasses among all these, the one that is characterized by the most distinct features, gets its inspiration from human cognitive and social behaviors such as learning, teaching, creativity, and teamwork. The present paper is a thorough review of human-inspired metaheuristic algorithms, and it is going to analyze their basic principles, types, and operational frameworks. The authors will be going into details about the mechanics of well-known algorithms such as Sewing Training-Based Optimization (STBO), Carpet Weaver Optimization (CWO), and the iHow Optimization Algorithm (iHowOA), emphasizing their individual methods for maintaining a balance between global exploration and local exploitation. To support the review, extensive comparative tables will summarize performance on standard benchmark functions and a broad range of real-world applications, including but not limited to, engineering design and feature selection, healthcare, and energy management. The quality of the algorithms deployed in this analysis is confirmed to be very good. They use structured human-like processes to effectively navigate through complex solution spaces. However, in line with the "No Free Lunch" theorem, their superiority is condition-based. The paper ends with a discussion of future research directions, emerging trends, and inherent challenges such as the potential for adaptive and hybrid models to further enhance robustness and versatility in dynamic optimization landscapes.