Lightly-supervised Clustering Using Pairwise Constraint Propagation

摘要：

This paper focuses on providing a high-quality semi-supervised clustering with small quantities of constraints.A Clustering method called CP-KMeans is proposed forpropagating pairwise constraints to nearby instancesusing a Gaussian function.This method takes a few easilyspecified constraints,and propagates them to nearbypairs of points to constrain the local neighborhood.Clustering with these propagated constraints can yieldsuperior performance with fewer constraints thanclustering with only the original user-specified constraints.The experimental results on several data sets show thatCP-KMeans obtain high performance with fewerconstraints compared with other two semi-supervisedclustering algorithms.

作者: Jianbin Huang Heli Sun

作者单位: School of Software,Xidian University Department of Computer Science and Technology,Xian Jiaotong Universit

会议类型: 国际会议

会议名称: 2008 3rd International Conference on Intelligent System and Knowledge Engineering(第三届智能系统与知识工程国际会议)(ISKE 2008)

会议地点: 厦门

会议语种:英文

页码: 765-770

在线出版日期: 2008-11-17（万方平台首次上网日期，不代表论文的发表时间）