File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Separable linear discriminant analysis

TitleSeparable linear discriminant analysis
Authors
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
Citation
Computational Statistics And Data Analysis, 2012, v. 56 n. 12, p. 4290-4300 How to Cite?
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.
Persistent Identifierhttp://hdl.handle.net/10722/172504
ISSN
2014 Impact Factor: 1.400
2013 SCImago Journal Rankings: 1.399
ISI Accession Number ID
References

 

DC FieldValueLanguage
dc.contributor.authorZhao, Jen_US
dc.contributor.authorYu, PLHen_US
dc.contributor.authorShi, Len_US
dc.contributor.authorLi, Sen_US
dc.date.accessioned2012-10-30T06:22:50Z-
dc.date.available2012-10-30T06:22:50Z-
dc.date.issued2012en_US
dc.identifier.citationComputational Statistics And Data Analysis, 2012, v. 56 n. 12, p. 4290-4300en_US
dc.identifier.issn0167-9473en_US
dc.identifier.urihttp://hdl.handle.net/10722/172504-
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.en_US
dc.languageengen_US
dc.publisherElsevier BV. The Journal's web site is located at http://www.elsevier.com/locate/csdaen_US
dc.relation.ispartofComputational Statistics and Data Analysisen_US
dc.subjectFace Recognitionen_US
dc.subjectLinear Discriminant Analysisen_US
dc.subjectSeparableen_US
dc.subjectTwo-Dimensional Dataen_US
dc.titleSeparable linear discriminant analysisen_US
dc.typeArticleen_US
dc.identifier.emailYu, PLH: plhyu@hku.hken_US
dc.identifier.authorityYu, PLH=rp00835en_US
dc.description.naturelink_to_subscribed_fulltexten_US
dc.identifier.doi10.1016/j.csda.2012.04.003en_US
dc.identifier.scopuseid_2-s2.0-84864148086en_US
dc.identifier.hkuros210593-
dc.relation.referenceshttp://www.scopus.com/mlt/select.url?eid=2-s2.0-84864148086&selection=ref&src=s&origin=recordpageen_US
dc.identifier.volume56en_US
dc.identifier.issue12en_US
dc.identifier.spage4290en_US
dc.identifier.epage4300en_US
dc.identifier.isiWOS:000307483100038-
dc.publisher.placeNetherlandsen_US
dc.identifier.scopusauthoridZhao, J=7410313775en_US
dc.identifier.scopusauthoridYu, PLH=7403599794en_US
dc.identifier.scopusauthoridShi, L=36078323000en_US
dc.identifier.scopusauthoridLi, S=55207163700en_US

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats