摘要
在K-Means聚类、PSO聚类、K-Means和PSO混合聚类(KPSO)的基础上,研究了基于量子行为的微粒群优化算法(QPSO)的数据聚类方法,并提出利用K-Means聚类的结果重新初始化粒子群,结合QPSO的聚类算法,即KQPSO。介绍了如何利用上述算法找到用户指定的聚类个数的聚类中心。聚类过程都是根据数据之间的Euclidean(欧几里得)距离。K-Means算法、PSO算法和QPSO算法的不同在于聚类中心向量的“进化”上。最后使用三个数据集比较了上面提到的五种聚类方法的性能,结果显示基于QPSO算法的数据聚类性能比一般PSO算法更好。
This paper investigates Quantum-behaved Particle Swarm Optimization (QPSO) algorithm to cluster data based on the K-Means clustering, PSO clustering and KPSO clustering. After that we introduce using K-Means clustering to seed the initial swarm, combing with QPSO to cluster data, namely KQPSO and introduce how these algorithms can be used to find the centroids of a user specified number of clusters. All the process of clustering based on the Euclidean distance among data vectors. The differences between K-Means, PSO, QPSO is the evolution of the cluster-centroids. Finally, we compare the performance of the five clustering method on three data sets. The experiments result show QPSO clustering superiority.
出处
《计算机应用研究》
CSCD
北大核心
2006年第12期40-42,45,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(60474030)
关键词
聚类
K—Means
PSO
QPSO
聚类中心
Clustering
K-Means
PSO
QPSO( Quantum-behaved Particle Swarm Optimization)
Cluster-centroids