Table 2.

BAS and related improvement strategies.

Improvement approach	Specific improvement strategies	Pros and cons	Related algorithms
Parameter adjustment	Adjusts step size, beetle spacing, introduces beetle populations.	Pros: Enhances global/local search. Cons: High complexity, sensitive to parameters.	VSBAS, BSAS, BASL, etc.
Adaptive mechanisms	Uses inertia weights, elite selection, fallback mechanisms.	Pros: Fast convergence, robust. Cons: Complexity, local optima risk.	BAS-ADAM, WSBAS, EBAS, ENBAS, FBAS, etc.
Hybrid heuristics	Combines PSO, ABC, FPA, GA, ACO, etc. for global/local search.	Pros: Combines strengths, versatile. Cons: High complexity, tuning needed.	BSO, BAS-PSO, BAPSO, MBAS, BAS-ABC, etc.
Deep learning	Optimizes neural networks (BP, CNN, ELM) with BAS.	Pros: Improves training speed/accuracy. Cons: High complexity, resource-heavy.	BASNNC, BASZNN, BAS-CNN, etc.