计算机工程与应用 ›› 2010, Vol. 46 ›› Issue (36): 168-170.DOI: 10.3778/j.issn.1002-8331.2010.36.046

• 数据库、信号与信息处理 • 上一篇    下一篇

近邻传播半监督聚类算法的分析与改进

赵宪佳1,王立宏2   

  1. 1.青岛大学 国际学院,山东 青岛 266071
    2.烟台大学 计算机学院,山东 烟台 264005
  • 收稿日期:2009-07-06 修回日期:2009-09-03 出版日期:2010-12-21 发布日期:2010-12-21
  • 通讯作者: 赵宪佳

Analysis and improvement of semi-supervised clustering algorithm based on affinity propagation

ZHAO Xian-jia1,WANG Li-hong2   

  1. 1.International College,Qingdao University,Qingdao,Shandong 266071,China
    2.School of Computer Science and Technology,Yantai University,Yantai,Shandong 264005,China
  • Received:2009-07-06 Revised:2009-09-03 Online:2010-12-21 Published:2010-12-21
  • Contact: ZHAO Xian-jia

摘要: 近邻传播半监督聚类算法SAP在小数据集上运行时可能会出现并列类代表点的现象,当出现并列类代表点时,依据决策矩阵E对角线上数值大于0确定的类代表点并不是全部的类代表点。分析了近邻传播算法的性质,找出了并列类代表点的出现原因,并针对此现象给出了改进算法。

关键词: 近邻传播, 类代表点, 半监督学习

Abstract: Parity exemplars often appear when Semi-supervised clustering algorithm based on Affinity Propagation(SAP) is applied on small dataset,and then the exemplar judgment criterion,i.e. an exemplar xk must satisfy E(k,k)>0 in the decision matrix E,is not complete.In this paper,properties of affinity propagation algorithm are analyzed and the occurrence reason of parity exemplars is found.Finally,an improved algorithm is proposed to solute this problem.

Key words: affinity propagation, exemplar, semi-supervised learning

中图分类号: