Linear dimensionality reduction using relevance weighted lda

EK Tang, PN Suganthan, Xin Yao, AK Qin

Research output: Contribution to journalArticle

77 Citations (Scopus)


The linear discriminant analysis (LDA) is one of the most traditional linear dimensionality reduction methods. This paper incorporates the inter-class relationships as relevance weights into the estimation of the over all within-class scatter matrix in order to improve the performance of the basic LDA method and some of its improved variants. We demonstrate that in son-re specific situations the standard multi-class LDA almost totally fails to find a discriminative sub-space if the proposed relevance weights are not incorporated. In order to estimate the relevance weights of individual within-class scatter matrices. we propose several methods of which one employs the evolution strategies. (C) 2004 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.
Original languageEnglish
Pages (from-to)485-493
Number of pages9
JournalPattern Recognition
Issue number4
Publication statusPublished - 1 Apr 2005


  • feature extraction
  • weighted LDA
  • Mahalanobis distance
  • evolution strategies
  • linear discriminant analysis
  • Chernoff criterion
  • approximate pairwise accuracy criterion


Dive into the research topics of 'Linear dimensionality reduction using relevance weighted lda'. Together they form a unique fingerprint.

Cite this