Article: Separable linear discriminant analysis

File Download Links for fulltext
(May Require Subscription)
Supplementary

  • Basic View
  • Metadata View
  • XML View
TitleSeparable linear discriminant analysis
AuthorsZhao, J2
Yu, PLH1
Shi, L2
Li, S2
KeywordsFace Recognition
Linear Discriminant Analysis
Separable
Two-Dimensional Data
Issue Date2012
PublisherElsevier BV. The Journal's web site is located at http://www.elsevier.com/locate/csda
CitationComputational Statistics And Data Analysis, 2012, v. 56 n. 12, p. 4290-4300 [How to Cite?]
DOI: http://dx.doi.org/10.1016/j.csda.2012.04.003
AbstractLinear discriminant analysis (LDA) is a popular technique for supervised dimension reduction. Due to the curse of dimensionality usually suffered by LDA when applied to 2D data, several two-dimensional LDA (2DLDA) methods have been proposed in recent years. Among which, the Y2DLDA method, introduced by Ye et al. (2005), is an important development. The idea is to utilize the underlying 2D data structure to seek for an optimal bilinear transformation. However, it is found that the proposed algorithm does not guarantee convergence. In this paper, we show that the utilization of a bilinear transformation for 2D data is equivalent to modeling the covariance matrix of 2D data as separable covariance matrix. Based on this result, we propose a novel 2DLDA method called separable LDA (SLDA). The main contributions of SLDA include (1) it provides interesting theoretical relationships between LDA and some 2DLDA methods; (2) SLDA provides a building block for mixture extension; (3) unlike Y2DLDA, a neatly analytical solution can be obtained as that in LDA. Empirical results show that our proposed SLDA achieves better recognition performance than Y2DLDA while being computationally much more efficient. © 2012 Elsevier B.V. All rights reserved.
ISSN0167-9473
2011 Impact Factor: 1.028
2011 SCImago Journal Rankings: 0.068
DOIhttp://dx.doi.org/10.1016/j.csda.2012.04.003
ReferencesReferences in Scopus
DC Field
Value
dc.contributor.authorZhao, J
dc.contributor.authorYu, PLH
dc.contributor.authorShi, L
dc.contributor.authorLi, S
dc.date.accessioned2012-10-30T06:22:50Z
dc.date.available2012-10-30T06:22:50Z
dc.date.issued2012
dc.description.abstractLinear discriminant analysis (LDA) is a popular technique for supervised dimension reduction. Due to the curse of dimensionality usually suffered by LDA when applied to 2D data, several two-dimensional LDA (2DLDA) methods have been proposed in recent years. Among which, the Y2DLDA method, introduced by Ye et al. (2005), is an important development. The idea is to utilize the underlying 2D data structure to seek for an optimal bilinear transformation. However, it is found that the proposed algorithm does not guarantee convergence. In this paper, we show that the utilization of a bilinear transformation for 2D data is equivalent to modeling the covariance matrix of 2D data as separable covariance matrix. Based on this result, we propose a novel 2DLDA method called separable LDA (SLDA). The main contributions of SLDA include (1) it provides interesting theoretical relationships between LDA and some 2DLDA methods; (2) SLDA provides a building block for mixture extension; (3) unlike Y2DLDA, a neatly analytical solution can be obtained as that in LDA. Empirical results show that our proposed SLDA achieves better recognition performance than Y2DLDA while being computationally much more efficient. © 2012 Elsevier B.V. All rights reserved.
dc.description.natureLink_to_subscribed_fulltext
dc.identifier.citationComputational Statistics And Data Analysis, 2012, v. 56 n. 12, p. 4290-4300 [How to Cite?]
DOI: http://dx.doi.org/10.1016/j.csda.2012.04.003
dc.identifier.doihttp://dx.doi.org/10.1016/j.csda.2012.04.003
dc.identifier.epage4300
dc.identifier.hkuros210593
dc.identifier.issn0167-9473
2011 Impact Factor: 1.028
2011 SCImago Journal Rankings: 0.068
dc.identifier.issue12
dc.identifier.scopuseid_2-s2.0-84864148086
dc.identifier.spage4290
dc.identifier.urihttp://hdl.handle.net/10722/172504
dc.identifier.volume56
dc.languageeng
dc.publisherElsevier BV. The Journal's web site is located at http://www.elsevier.com/locate/csda
dc.publisher.placeNetherlands
dc.relation.ispartofComputational Statistics and Data Analysis
dc.relation.referencesReferences in Scopus
dc.subjectFace Recognition
dc.subjectLinear Discriminant Analysis
dc.subjectSeparable
dc.subjectTwo-Dimensional Data
dc.titleSeparable linear discriminant analysis
dc.typeArticle
Author Affiliations
  1. The University of Hong Kong
  2. Yunnan University