Metaheuristics for clustering problems

Data clustering is an essential, unsupervised classification method for discovering hidden patterns or information of a dataset. The literature shows metaheuristics are a reliable alternative to conventional clustering algorithms for data clustering problems. In this chapter, we compare 12 metaheuristics in this area. The algorithms used are Genetic Algorithm (GA), Gray Wolf Optimizer (GWO), Differential Evolution (DE), Biogeography-based Optimization (BBO), Harmony Search (HS), Particle Swarm Optimization (PSO), African Vulture Optimization Algorithm (AVOA), Firefly Algorithm (FFA), Symbiotic Organism Search (SOS), Artificial Bee Colony (ABC) algorithm, Whale Optimization Algorithm (WOA), and Artificial Gorilla Troops Optimization (AGTO). These algorithms were selected based on their unique capabilities. To test and evaluate the performance of the selected algorithms to solve the clustering problem, we use 10 standard datasets in the UCI repository. In addition, we use the ANOVA statistical test to examine the significant differences in the solutions obtained from the compared optimization algorithms.