View Item 

The matrix form for weighted linear discriminant analysis and fractional linear discriminant analysis

Access Status
Open access
Authors
Xu, T.
Lu, C.
Liu, Wan-quan
Date
2009
Type
Conference Paper

Metadata
Show full item record
Citation
Xu, T. and Lu, C. and Liu, Wan-quan. 2009. The matrix form for weighted linear discriminant analysis and fractional linear discriminant analysis, in Lai, L. L. (ed), ICMLC 2009, Jul 12 2009, pp. 1621-1627.Baoding, Hebei, China: IEEE.
Source Title
Proceedings of the 8th International conference on Machine learning and Cybernetics
Source Conference
ICMLC 2009
Additional URLs
ISBN
Faculty
School of Science and Computing
Department of Computing
Faculty of Science and Engineering
Remarks

Copyright © 2009 IEEE This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

URI
Collection
Abstract

In this paper we will extend the recently proposed weighted linear discriminant analysis (W_LDA) and fraction-step linear discriminant analysis (F_LDA) from one dimension vector form to the case of two dimension matrix form, which are called weighted two dimensional linear discriminant analysis (W_2DLDA) and fraction-step two dimension linear discriminant analysis (F_2DLDA), respectively. The motivation of this work is based on the recent research results on two dimensional principal component analysis (2DPCA) and 2DLDA showing that the two dimensional algorithms can save computational costs significantly and thus improve the classifiers performances. First, we derived these numerical algorithms in matrix form and then we implement these two new algorithms on ORL and YALE face databases. The experimentation results show that W_2DLDA produces the best performance among F_2DLDA, F_LDA and W_LDA.