Clustering analysis using Swarm Intelligence

This thesis is concerned with the application of the swarm intelligence methods in clustering analysis of datasets. The main objectives of the thesis are ∙ Take the advantage of a novel evolutionary algorithm, called artificial bee colony, to improve the capability of K-means in finding global optimum clusters in nonlinear partitional clustering problems. ∙ Consider partitional clustering as an optimization problem and an improved antbased algorithm, named Opposition-Based API (after the name of Pachycondyla APIcalis ants), to automatic grouping of large unlabeled datasets. ∙ Define partitional clustering as a multiobjective optimization problem. The aim is to obtain well-separated, connected, and compact clusters and for this purpose, two objective functions have been defined based on the concepts of data connectivity and cohesion. These functions are the core of an efficient multiobjective particle swarm optimization algorithm, which has been devised for and applied to automatic grouping of large unlabeled datasets. For that purpose, this thesis is divided is five main parts: ∙ The first part, including Chapter 1, aims at introducing state of the art of swarm intelligence based clustering methods. ∙ The second part, including Chapter 2, consists in clustering analysis with combination of artificial bee colony algorithm and K-means technique. ∙ The third part, including Chapter 3, consists in a presentation of clustering analysis using opposition-based API algorithm. ∙ The fourth part, including Chapter 4, consists in multiobjective clustering analysis using particle swarm optimization. ∙ Finally, the fifth part, including Chapter 5, concludes the thesis and addresses the future directions and the open issues of this research.