Data Clustering Using Chaotic Particle Swarm Optimization.docVIP

IAENG International Journal of Computer Science, 39:2, IJCS_39_2_08 ______________________________________________________________________________________ Data Clustering Using Chaotic Particle Swarm Optimization Li-Yeh Chuang, Yu-Da Lin, and Cheng-Hong Yang, Member, IAENG ? methods such as genetic algorithms [3] and particle swarm optimization [4-6]. Genetic algorithms typically start with some candidate solutions to an optimization problem. These candidates then evolve towards a better solution through selection, crossover and mutation. The concept of PSO was designed to simulate social behavior via an exchange of information in practical applications. Many studies used PSO to cluster data within a multi-dimensional space and obtained outstanding results. However, the rate of convergence is insufficient when searching for global optima. Fan et al., [7] proposed to combine the Nelder–Mead simplex search method with PSO to improve it. The rationale behind it was that such a hybrid approach enjoys the merits of both PSO and the Nelder–Mead simplex search method. Kao et al. explored the applicability of a hybrid K-means algorithm, Nelder-Mead simplex search method, and particle swarm optimization (K–NM–PSO) to clustering data vectors [8]. PSO adopts a random sequence with a random starting point as a parameter, and relies on this parameter to update the positions and velocities of the particles. However, PSO often leads to premature convergence, especially in complex multi-peak search problems such as the clustering of high-dimensional data. We combined a Gauss chaotic map and particle swarm optimization in a new method named GaussPSO. Results of the conducted experimental trials on a variety of data sets taken from several real-life situations demonstrate that the proposed GaussPSO method is superior to the K-means, PSO, NM-PSO, K-PSO, and K-NM-PSO a