摘要
针对现有聚类算法在参数输入、停机条件等方面存在诸多人为控制因素的问题,采用信息熵理论使聚类标准客观化,同时结合模糊聚类的思想,以隶属度作为信息熵计算的基础,并采用谱系的方法确定聚类数目,从而改进模糊聚类算法.研究表明,提出的基于信息熵的算法能够比较客观、科学地反映实际聚类情况.
To answer the questions of the existing clustering algorithms involving many man-made factors such as parameters input and the pausing condition of clustering algorithm, this paper objectified the clustering standard by the entropy information theory, and put forward a new kind of clustering algorithm to avoid the deficiency based on the theory fuzzy set and information entropy. It is shown that the algorithm based on entropy of information put forward in this paper can reflect the real cluster′s situation objectively and scientifically.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2005年第2期92-94,共3页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
关键词
聚类算法
信息熵
模糊聚类
谱系方法
clustering algorithm
information entropy
fuzzy clustering
pedigree clustering
