计算机应用 ›› 2012, Vol. 32 ›› Issue (03): 646-648.DOI: 10.3724/SP.J.1087.2012.00646

• 人工智能 • 上一篇    下一篇

距离修正的模糊C均值聚类算法

楼晓俊1,李隽颖1,刘海涛1,2   

  1. 1.中国科学院 上海微系统与信息技术研究所,上海 200050;
    2.无锡物联网产业研究院,江苏 无锡 214135
  • 收稿日期:2011-08-22 修回日期:2011-12-08 发布日期:2012-03-01 出版日期:2012-03-01
  • 通讯作者: 楼晓俊
  • 作者简介:楼晓俊(1984-),男,浙江杭州人,博士研究生,CCF会员,主要研究方向:传感器网络信号处理、模式识别;李隽颖(1982-),男,湖北云梦人,博士研究生,CCF会员,主要研究方向:传感器网络信号处理、模式识别;刘海涛(1968-),男,新疆昌吉人,研究员,博士生导师,主要研究方向:传感器网络、物联网体系架构。
  • 基金资助:

    国家科技重大专项(2010ZX03006-004);国家973计划项目(2011CB302906)。

Improved fuzzy C-means clustering algorithm based on distance correction

LOU Xiao-jun1, LI Jun-ying1, LIU Hai-tao1,2   

  1. 1.Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, Shanghai 200050, China;
    2.Wuxi SensingNet Industrialization Research Institute, Wuxi Jiangsu 214135,China
  • Received:2011-08-22 Revised:2011-12-08 Online:2012-03-01 Published:2012-03-01
  • Contact: Xiao-Jun LOU

摘要: 经典的模糊C均值算法基于欧氏距离,存在等划分趋势的缺陷,分错率较高,只适用于球形结构的聚类。针对这一问题,利用数据的点密度信息,在数据点与聚类中心的距离度量中引入了调节因子,提出了一种基于密度的距离修正矩阵,并用其代替经典模糊C均值算法中的距离度量矩阵。通过人造数据集和UCI数据集的两组聚类实验,证实了改进算法对非球形结构的数据同样适用,且相比经典的模糊C均值算法具有更高的聚类准确率。

关键词: 聚类, 模糊C均值, 距离度量, 点密度, 调节因子

Abstract: Based on Euclidean distance, the classic Fuzzy C-Means (FCM) clustering algorithm has the limitation of equal partition trend for data sets. And the clustering accuracy is lower when the distribution of data points is not spherical. To solve these problems, a distance correction factor based on dot density was introduced. Then a distance matrix with this factor was built for measuring the differences between data points. Finally, the new matrix was applied to modify the classic FCM algorithm. Two sets of experiments using artificial data and UCI data were operated, and the results show that the proposed algorithm is suitable for non-spherical data sets and outperforms the classic FCM algorithm in clustering accuracy.

Key words: clustering, Fuzzy C-Means (FCM), distance measurement, dot density, regulatory factor

中图分类号: