On the Required Iteration Number of the Representant Swarm Intelligence Algorithms Based on Coefficient of Variation

Determination of the required parameters for different optimization algorithms used in specific environments was and remains a hot topic of the informatics sciences. Selection of the right tool for the problem set is hard and careful investigation is paramount. This scientific study investigates the effectiveness of swarm intelligence-based optimization methods, namely Artificial Bee Colony, Grey Wolf optimizer, Multiverse optimizer and Salp Swarm Algorithm methods on four well known search space functions: Sphere, Rosenbrock’s Saddle, Rastring’s function and Ackley’s function. Coefficient of variation of the cost functions was used to evaluate quantitatively the performance of the analysed searching spaces. Ranking of the algorithms was done for each of the optimization problem wherein the SSA algorithm demonstrated a higher degree of favourability across the majority of the search spaces investigated.