CHINESE JOURNAL OF APPLIED PROBABILITY AND STATIST 2008, 24(5) 475-483 DOI:      ISSN: 1001-4268 CN: 31-1256

Current Issue | Archive | Search                                                            [Print]   [Close]
ѧ������
Information and Service
This Article
Supporting info
PDF(500KB)
[HTML]
Reference
Service and feedback
Email this article to a colleague
Add to Bookshelf
Add to Citation Manager
Cite This Article
Email Alert
Keywords
Gaussian mixture models
unsupervised Classification
penalized maximum likelihood
EM algorithm
invert Wishart distribution.
Authors
Yu Peng
Tong Xinwei
Feng Jufu
PubMed
Article by
Article by
Article by

�������ͷ���Ȼ�ĸ�˹���ģ���޼ල�����о�

Yu Peng, Tong Xinwei, Feng Jufu

School of Mathematical Sciences, Peking University, National Geomatics Center of China; School of Mathematical Sciences, Beijing Normal University; National Laboratory on Machine Perception, Center for Information Science, School of Electronics Engineering and Computer Science, Peking University

Abstract��

In this paper we propose an unsupervised classification algorithm which is based on Gaussian mixture models. Thinking that EM algorithm will result in a local optimal resolution of Gaussian mixture models in parameter estimations, we substitute invert Wishart distribution for Jeffery prior. Experiments show that this
algorithm improves correct rates and decreases time while estimating
classifications.

Keywords�� Gaussian mixture models   unsupervised Classification   penalized maximum likelihood   EM algorithm   invert Wishart distribution.  
Received 1900-01-01 Revised 1900-01-01 Online:  
DOI:
Fund:
Corresponding Authors: Yu Peng
Email:
About author:

References��
Similar articles

Copyright by CHINESE JOURNAL OF APPLIED PROBABILITY AND STATIST