. 2025 Feb 28;11:e2722. doi: 10.7717/peerj-cs.2722

Table 2. Overall comparison between ASO and state-of-the-art algorithms.

Ref.	Algorithm	Year	Category	Parameter	Complexity	Merits	Limitations
Zhao, Wang & Zhang (2019b)	Atom search optimization (ASO)	2019	Physics-based	2	$O (T N D + T C N)$	Easy implementation and high Capability	Lower population diversity and may struggle with high-dimensional problems
Kennedy & Eberhart (1995)	Particle swarm algorithm (PSO)	1995	Swarm-based	4	$O (N + N \times T)$	Simple to implement, fast convergence	Prone to premature convergence, sensitive to parameter tuning
Mirjalili & Lewis (2016)	Whale optimization algorithm (WOA)	2016	Swarm-based	5	$O (T N D + T N)$	Simple implementation, good exploration	Probability of falling into local optima
Wang et al. (2022a)	Artificial rabbit optimizer (ARO)	2022	Swarm-based	2	$O (T N D + T N + N)$	Flexibility, stability, good balance between exploration and exploitation	Slow convergence
Faramarzi et al. (2020)	Marine predators algorithm (MPA)	2020	Swarm-based	3	$O (T N D + T N)$	Effective in handling multimodal problems	Limited in global search capability
Heidari et al. (2019)	Harris Hawks optimization (HHO)	2019	Swarm-based	2	$O (T N D + T N + N)$	Strong exploitation ability	Convergence may slow in high dimensions
Gandomi, Yang & Alavi (2013)	Cuckoo search (CS)	2009	Swarm-based	1	$O (N + 2 N T)$	Performs well in complex problems	Suffer from slow convergence
Mirjalili, Mirjalili & Lewis (2014)	Grey wolf optimizer (GWO)	2014	Swarm-based	1	$O (N + N \times T)$	Simple structure, easy to implement	Limited exploration capabilities
Holland (1992)	Genetic algorithm (GA)	1992	Evolution-based	3	$O (T N + N)$	Handles discrete and continuous optimization	Evaluation is relatively expensive
Storn & Price (1997)	Differential evolution (DE)	1997	Evolution-based	2	$O (T N + N)$	Effective for continuous optimization	May require many function evaluations
Geem, Kim & Loganathan (2001)	Harmony search (HS)	2001	Human-based	3	$O (N + N \times T)$	Suitable for discrete and continuous problems	May require parameter tuning
Van Laarhoven et al. (1987)	Simulated annealing (SA)	1987	Physics-based	2	$O (1 + T)$	Good global optimization ability	Slow convergence in large search spaces
Mirjalili (2016)	Sine cosine algorithm (SCA)	2016	Mathematics-based	1	$O (T N D + T N + N)$	Simple, few parameters, balances exploration and exploitation	May get stuck in local optima