Article: Separable linear discriminant analysis
| Title | Separable linear discriminant analysis |
|---|---|
| Authors | Zhao, J2 Yu, PLH1 Shi, L2 Li, S2 |
| Keywords | Face Recognition Linear Discriminant Analysis Separable Two-Dimensional Data |
| Issue Date | 2012 |
| Publisher | Elsevier 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?] DOI: http://dx.doi.org/10.1016/j.csda.2012.04.003 |
| Abstract | Linear 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. |
| ISSN | 0167-9473 2011 Impact Factor: 1.028 2011 SCImago Journal Rankings: 0.068 |
| DOI | http://dx.doi.org/10.1016/j.csda.2012.04.003 |
| References | References in Scopus |
| dc.contributor.author | Zhao, J |
|---|---|
| dc.contributor.author | Yu, PLH |
| dc.contributor.author | Shi, L |
| dc.contributor.author | Li, S |
| dc.date.accessioned | 2012-10-30T06:22:50Z |
| dc.date.available | 2012-10-30T06:22:50Z |
| dc.date.issued | 2012 |
| dc.description.abstract | Linear 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.nature | Link_to_subscribed_fulltext |
| dc.identifier.citation | Computational 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.doi | http://dx.doi.org/10.1016/j.csda.2012.04.003 |
| dc.identifier.epage | 4300 |
| dc.identifier.hkuros | 210593 |
| dc.identifier.issn | 0167-9473 2011 Impact Factor: 1.028 2011 SCImago Journal Rankings: 0.068 |
| dc.identifier.issue | 12 |
| dc.identifier.scopus | eid_2-s2.0-84864148086 |
| dc.identifier.spage | 4290 |
| dc.identifier.uri | http://hdl.handle.net/10722/172504 |
| dc.identifier.volume | 56 |
| dc.language | eng |
| dc.publisher | Elsevier BV. The Journal's web site is located at http://www.elsevier.com/locate/csda |
| dc.publisher.place | Netherlands |
| dc.relation.ispartof | Computational Statistics and Data Analysis |
| dc.relation.references | References in Scopus |
| dc.subject | Face Recognition |
| dc.subject | Linear Discriminant Analysis |
| dc.subject | Separable |
| dc.subject | Two-Dimensional Data |
| dc.title | Separable linear discriminant analysis |
| dc.type | Article |
Author Affiliations
- The University of Hong Kong
- Yunnan University

