计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (11): 142-144.

• 数据库、信号与信息处理 • 上一篇    下一篇

聚类组合研究的新进展

杨 燕1,靳 蕃1,KAMEL Mohamed2   

  1. 1.西南交通大学 信息科学与技术学院,成都 610031
    2.加拿大滑铁卢大学 电子与计算机工程系,安大略 滑铁卢 N2L 3G1
  • 收稿日期:2007-09-29 修回日期:2007-11-12 出版日期:2008-04-11 发布日期:2008-04-11
  • 通讯作者: 杨 燕

Latest development of clustering ensemble

YANG Yan1,JIN Fan1,KAMEL Mohamed2   

  1. 1.School of Information Science and Technology,Southwest Jiaotong University,Chengdu 610031,China
    2.Department of Electrical and Computing Engineering,University of Waterloo,Waterloo N2L 3G1,Canada
  • Received:2007-09-29 Revised:2007-11-12 Online:2008-04-11 Published:2008-04-11
  • Contact: YANG Yan

摘要: 作为目前聚类分析的新兴研究热点,聚类组合方法能将两种或多种聚类方法集成起来以改善其性能。从聚类多样性和共识函数两方面综述了最新研究进展,探讨将神经网络组合的思想用于聚类组合。最后指出了将来可能的研究方向。

关键词: 聚类组合, 多样性分量, 共识函数, 神经网络组合

Abstract: As a novel research hotspot of clustering analysis currently,clustering ensemble can improve the performance of data clustering by combining two or multiple clustering algorithms.Some latest research results on clustering diversity and consensus function are reviewed,and a clustering ensemble method inspired by neural network ensemble is presented.Finally the future research issues are discussed.

Key words: clustering ensemble, diversity component, consensus function, neural network ensemble