On applying linear discriminant analysis for multi-labeled problems

被引：40

作者：

Park, Cheong Hee ^{[1
]}

Lee, Moonhwi ^{[1
]}

机构：

[1] Chungnam Natl Univ, Dept Comp Sci & Engn, Taejon 305763, South Korea

来源：

PATTERN RECOGNITION LETTERS | 2008年 / 29卷 / 07期

基金：

新加坡国家研究基金会;

关键词：

dimension reduction; linear discriminant analysis; multi-labeled problems; text categorization;

D O I：

10.1016/j.patrec.2008.01.003

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Linear discriminant analysis (LDA) is one of the most popular dimension reduction methods, but it is originally focused on a single-abeled problem. In this paper, we derive the formulation for applying LDA for a multi-labeled problem. We also propose a generalized LDA algorithm which is effective in a high dimensional multi-labeled problem. Experimental results demonstrate that by considering multi-labeled structure, LDA can achieve computational efficiency and also improve classification performances. (C) 2008 Elsevier B.V. All rights reserved.

引用

页码：878 / 887

页数：10

共 22 条

[1]

BELLHUMEUR P, 1997, IEEE T PATTERN ANAL, V19, P711

[2] A new LDA-based face recognition system which can solve the small sample size problem [J].