计算机与现代化

• 算法设计与分析 • 上一篇    下一篇

改进的基于熵的中心聚类算法

  

  1. 山东理工大学计算机系,山东淄博255049
  • 收稿日期:2013-11-05 出版日期:2014-03-24 发布日期:2014-03-31
  • 作者简介:张树森(1988-),男,山东东平人,山东理工大学计算机系硕士研究生,研究方向:数据挖掘算法; 通讯作者:张龙波(1968-),男,教授,博士,研究方向:数据库理论与技术。
  • 基金资助:
    山东省自然科学基金资助项目(ZR2011FL013); 山东省高等学校科技计划项目(J13LN27)

Improved Entropy-based Centre Clustering Algorithm

  1. Department of Computer Science, Shandong University of Science and Technology, Zibo 255049, China
  • Received:2013-11-05 Online:2014-03-24 Published:2014-03-31

摘要: 依据基于熵的模糊聚类算法(EFC),提出一种改进的基于熵的中心聚类算法,即通过EFC算法得到差异性十分明显的原始数据集的簇心,以这些簇心为中心再次进行聚类分析,通过各点到各中心的距离将各点重新分配到以各中心所代表的集合中。改进的算法不仅可以得到具有紧凑且差异明显的聚类结果,还可以使准确率得到有效提高。实验结果表明,该改进的算法能够实现数据集的有效聚类,相比于EFC算法的聚类结果准确率更高。

关键词: 聚类分析, EFC算法, 簇心

Abstract:  This article is based on entropy-based fuzzy clustering algorithm (EFC), proposes an improved entropy-based centre clustering algorithm, namely the very obvious differences clusters heart in the original data is obtained through EFC algorithm and these clusters heart to be heart-centered cluster is analyzed again, each point is re-assigned to the collection which the center represents through computing the distance between each point and center. This improved algorithm can not only get compact and significantly different clustering results, but also can effectively improve the accuracy rate. Experimental results show that the improved algorithm can realize data sets efficient clustering, with higher accuracy.

Key words: cluster analysis, EFC algorithm, cluster center

中图分类号: