Clustering Analysis Based on Chaos Immune Algorithm

Abstract

To improve the accuracy of clustering classification, the Chaos Immune Algorithm was proposed. In this algorithm, the ergodic property of chaos phenomenon is used to optimize the initial population, so it can accelerate the convergence of Immune Algorithms. Chaotic systems are sensitive to initial condition system parameters. Through the clone selection operator, antibody circulation and supplement, Clone operator and excellent individual chaotic disturbance, local optimums were avoided, so the global optimization was obtained. As for this issue, chaotic immune algorithm, ALECO-2, BPMA and BP algorithm were applied to test the algorithm performance in two group experiments. Theory and experiment showed that the Chaos Immune Algorithm can get global optimum clustering center, and greatly improve the amplitude of operation.

DOI : http://dx.doi.org/10.11591/telkomnika.v12i3.4426