计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (19): 177-180.DOI: 10.3778/j.issn.1002-8331.2009.19.055

• 图形、图像、模式识别 • 上一篇    下一篇

一种具有影响力因子的硬聚类算法

王建锋1,金 健2,王晶晶3   

  1. 1.上海电力学院 计算机科学与技术系,上海 200090
    2.华东师范大学 计算机科学与技术系,上海 200062
    3.上海西渡学校,上海 201401
  • 收稿日期:2008-04-21 修回日期:2008-07-23 出版日期:2009-07-01 发布日期:2009-07-01
  • 通讯作者: 王建锋

Study on influence of effectiveness factor in HCM algorithm

WANG Jian-feng1,JIN Jian2,WANG Jing-jing3   

  1. 1.Department of Computer Science and Technology,Shanghai University of Electric Power,Shanghai 200090,China
    2.Department of Computer Science and Technology,East China Normal University,Shanghai 200062,China
    3.Shanghai Xidu School,Shanghai 201401,China
  • Received:2008-04-21 Revised:2008-07-23 Online:2009-07-01 Published:2009-07-01
  • Contact: WANG Jian-feng

摘要: 为解决传统聚类方法对不同规模类不能正确聚类的问题,探讨了带影响力因子的硬聚类方法。为每个类均赋予一个影响力因子,使样本的隶属关系不只受距离的影响,而且受类的规模的影响。通过对18个数据集的实验,证明该方法的可行性,并且观察了影响力因子的取值对收敛过程和算法产生结果的影响,提出了今后的工作重点。

关键词: 聚类, 硬聚类算法, 模糊C均值算法, 影响力因子

Abstract: Focusing on the equalization in cluster size using traditional clustering method,Hard C-Means(HCM) algorithm with effectiveness factors is discussed.Assigning an effectiveness factor for each of the clusters,leading to the membership relations between samples affected not only by distances,but scales.Testing on 18 datasets verifies the feasibility of this method.Effects on convergence process and algorithm results caused by the effectiveness factor are observed,and the future emphasis is suggested.

Key words: clustering, Hard C-Means(HCM), Fuzzy C-Means(FCM), effectiveness factor