基于SOM和PSO的聚类组合算法

自组织映射（SOM）算法和粒子群优化（PSO）算法拥有概念简明、实现方便、收敛速度快、参数设置少，自组织自适应的特点。为了提高聚类效果，提出一种基于SOM和PSO的聚类组合算法SOM／PSO算法。先用SOM算法对UCI数据集中的数据进行聚类，然后以SOM的输出权值初始化PSO的粒子位置，再用PSO聚类算法进行聚类。实验结果表明所提出算法能够改进聚类性能，有很好的全局收敛性，不易陷入局部最优，也不会出现空簇现象。

Cluster Algorithm Based on Hybrid SOM and PSO

Recent years,people have paid much attention to the self-organizing feature maps(SOM) algorithm and the particle swarm optimization(PSO) algorithm in the clustering analysis field.These two algorithms both have the characters of concise conception,easy implementation,fast convergence,less parameters,self-organizing and self-adaptation.In order to improve the clustering results,a hybrid clustering algorithm named SOM/ PSO is proposed,which is based on SOM and PSO.Firstly the SOM algorithm is used to cluster ...