Survey on particle swarm optimization based clustering analysis
SIDE'12 Proceedings of the 2012 international conference on Swarm and Evolutionary Computation
Hi-index | 0.00 |
In order to solve the problem of easily fall into local optimal solutions, lower convergent precision, slower convergence rates and the poor population diversity, an improved PSO algorithm was proposed in this paper. The diversity was improved by the application of fuzzy clustering method. The sub-populations were classified automatically based on the feature of the population, and the information was exchanged by alliance in among the sub-populations. The simulation results of our improved PSO and indicated that the performance of optimal precision, efficiency and the stability are much better than that of traditional PSO.