计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (1): 173-173.

• 数据库与信息处理 • 上一篇    下一篇

基于辅助空间与极大熵的半监督聚类方法

罗晓清,王士同   

  1. 江南大学
  • 收稿日期:2006-01-04 修回日期:1900-01-01 出版日期:2007-01-01 发布日期:2007-01-01
  • 通讯作者: 罗晓清 zdlxq

Semi-supervised Clustering Based on Auxiliary Space and Maximum Entropy

,   

  1. 江南大学
  • Received:2006-01-04 Revised:1900-01-01 Online:2007-01-01 Published:2007-01-01

摘要: 摘 要 :将极大熵原理引入半监督聚类方法中,提出基于辅助空间与极大熵的半监督聚类算法AMESC,针对该算法中的代价函数进行迭代优化,实现聚类。AMESC的优势在于它依据模拟退火过程,使算法避开局部极小而得到全局极小,提高算法性能。通过实验证实了AMESC的有效性和优越性。

Abstract: Abstract:In this paper, the maximum-entropy approach is introduced to semi-supervised clustering, and a novel semi-supervised clustering algorithm AMESC based on auxiliary space, maximum entropy and simulated annealing is proposed. This algorithm realizes the efficient clustering by minimizing the cost function iteratively. Our experimental results demonstrate its validity and superiority.