Empirical Studies on Application of Genetic Algorithms and Ant Colony Optimization for Data Clustering

Cluster analysis is used in several research areas to classify data sets in groups by their similar characteristics. Metaheuristic-based techniques, such as Genetic Algorithms (GAs) and Ant Colony Optimization (ACO), have been applied in order to increase the clustering algorithm performance. GA and ACO-based clustering algorithms are capable of efficiently and automatically forming natural groups from a pre-defined number of clusters. This paper presents a GA and an ACO algorithm to the clustering problem. Both algorithms were refined using local search in order to improve the clustering accuracy. The results are compared on numeric UCI databases.