摘要
总结评述了K-means聚类算法的研究现状,指出K-means聚类算法是一个NP难优化问题,无法获得全局最优。介绍了K-means聚类算法的目标函数、算法流程,并列举了一个实例,指出了数据子集的数目K、初始聚类中心选取、相似性度量和距离矩阵为K-means聚类算法的3个基本参数。总结了K-means聚类算法存在的问题及其改进算法,指出了K-means聚类的进一步研究方向。
K-means clustering algorithm is reviewed.K-means clustering algorithm is a NP hard optimal problem and global optimal result cannot be reached.The goal,main steps and example of K-means clustering algorithm are introduced.K-means algorithm requires three user-specified parameters：number of clusters K,cluster initialization,and distance metric.Problems and improvement of K-means clustering algorithm are summarized then.Further study directions of K-means clustering algorithm are pointed at last.
出处
《电子设计工程》
2012年第7期21-24,共4页
Electronic Design Engineering
基金
国家自然科学基金资助项目(10776026)
