计算机工程与应用 ›› 2010, Vol. 46 ›› Issue (18): 132-134.DOI: 10.3778/j.issn.1002-8331.2010.18.042

• 数据库、信号与信息处理 • 上一篇    下一篇

一种混合均值聚类算法的实现

陈寿文1,2,李明东1   

  1. 1.西华师范大学 微机应用研究所,四川 南充 637000
    2.滁州学院 数学系,安徽 滁州 239000
  • 收稿日期:2008-12-08 修回日期:2009-03-23 出版日期:2010-06-21 发布日期:2010-06-21
  • 通讯作者: 陈寿文

Design of hybrid means algorithm for clustering

CHEN Shou-wen1,2,LI Ming-dong1   

  1. 1.Institute of Computer Application,China West Normal University,Nanchong,Sichuan 637000,China
    2.Department of Mathematics,Chuzhou University,Chuzhou,Anhui 239000,China
  • Received:2008-12-08 Revised:2009-03-23 Online:2010-06-21 Published:2010-06-21
  • Contact: CHEN Shou-wen

摘要: K-Means聚类算法和FCM算法混合运行的角度来探讨聚类问题,针对FCM算法初始化隶属度矩阵的随机性问题,提出了一种混合均值聚类算法。在混合算法运行过程中,利用前者的聚类结果信息来初始化后者的初始中心,依此来计算FCM算法初始隶属度矩阵,通过FCM算法的运行,最终实现数据集的聚类目的。实验结果表明该混合均值算法比单纯使用FCM算法效果好。

关键词: K-Means算法, 模糊C均值算法, 混合均值算法

Abstract: Based on the K-Means and Fuzzy C Means(FCM) algorithm’s mixing operation,this paper discusses the clustering problem and provides a hybrid means algorithm for clustering to solve the algorithm’s initial problem about the FCM.During the new algorithm’s operation,it uses the former’s result of clustering and evaluates that to the latter algorithm’s relevant variables.After that,the FCM algorithm computes its initial degree membership matrix,then,the FCM can function favorably till its end and fulfill the purpose of clustering the initial data sets.The experimental result indicates the hybrid algorithm effects nicely than the FCM does.

Key words: K-means algorithm, Fuzzy C Mean(FCM) algorithm, hybrid means algorithm

中图分类号: