DocumentCode :
1221443
Title :
Class conditional density estimation using mixtures with constrained component sharing
Author :
Titsias, Michalis K. ; Likas, Aristidis
Author_Institution :
Dept. of Comput. Sci., Ioannina Univ., Greece
Volume :
25
Issue :
7
fYear :
2003
fDate :
7/1/2003 12:00:00 AM
Firstpage :
924
Lastpage :
928
Abstract :
We propose a generative mixture model classifier that allows for the class conditional densities to be represented by mixtures having certain subsets of their components shared or common among classes. We argue that, when the total number of mixture components is kept fixed, the most efficient classification model is obtained by appropriately determining the sharing of components among class conditional densities. In order to discover such an efficient model, a training method is derived based on the EM algorithm that automatically adjusts component sharing. We provide experimental results with good classification performance.
Keywords :
pattern classification; pattern recognition; classifier; component sharing; conditional densities; density estimation; mixture model; Partitioning algorithms;
fLanguage :
English
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
Publisher :
ieee
ISSN :
0162-8828
Type :
jour
DOI :
10.1109/TPAMI.2003.1206521
Filename :
1206521
Link To Document :
بازگشت