L1-norm-based 2DPCA

Xuelong Li; Yanwei Pang; Yuan Yuan

doi:10.1109/TSMCB.2009.2035629

L1-norm-based 2DPCA

Xuelong Li^*, Yanwei Pang, Yuan Yuan

^*Corresponding author for this work

Computer Science Research Group

Research output: Contribution to journal › Article › peer-review

293 Citations (SciVal)

Abstract

In this paper, we first present a simple but effective L1-norm-based two-dimensional principal component analysis (2DPCA). Traditional L2-norm-based least squares criterion is sensitive to outliers, while the newly proposed L1-norm 2DPCA is robust. Experimental results demonstrate its advantages.

Original language	English
Pages (from-to)	1170-1175
Number of pages	6
Journal	IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Volume	40
Issue number	4
Early online date	15 Jan 2010
DOIs	https://doi.org/10.1109/TSMCB.2009.2035629
Publication status	Published - Aug 2010

Keywords

L1 norm
outlier
subspace
two-dimensional principal component analysis (2DPCA)

Access to Document

10.1109/TSMCB.2009.2035629

Cite this

@article{76469a97ceeb4358a83832798ad2de97,

title = "L1-norm-based 2DPCA",

abstract = "In this paper, we first present a simple but effective L1-norm-based two-dimensional principal component analysis (2DPCA). Traditional L2-norm-based least squares criterion is sensitive to outliers, while the newly proposed L1-norm 2DPCA is robust. Experimental results demonstrate its advantages.",

keywords = "L1 norm, outlier, subspace, two-dimensional principal component analysis (2DPCA)",

author = "Xuelong Li and Yanwei Pang and Yuan Yuan",

year = "2010",

month = aug,

doi = "10.1109/TSMCB.2009.2035629",

language = "English",

volume = "40",

pages = "1170--1175",

journal = "IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics",

issn = "1083-4419",

publisher = "IEEE",

number = "4",

}

TY - JOUR

T1 - L1-norm-based 2DPCA

AU - Li, Xuelong

AU - Pang, Yanwei

AU - Yuan, Yuan

PY - 2010/8

Y1 - 2010/8

N2 - In this paper, we first present a simple but effective L1-norm-based two-dimensional principal component analysis (2DPCA). Traditional L2-norm-based least squares criterion is sensitive to outliers, while the newly proposed L1-norm 2DPCA is robust. Experimental results demonstrate its advantages.

AB - In this paper, we first present a simple but effective L1-norm-based two-dimensional principal component analysis (2DPCA). Traditional L2-norm-based least squares criterion is sensitive to outliers, while the newly proposed L1-norm 2DPCA is robust. Experimental results demonstrate its advantages.

KW - L1 norm

KW - outlier

KW - subspace

KW - two-dimensional principal component analysis (2DPCA)

UR - http://www.scopus.com/inward/record.url?scp=77954762644&partnerID=8YFLogxK

U2 - 10.1109/TSMCB.2009.2035629

DO - 10.1109/TSMCB.2009.2035629

M3 - Article

C2 - 20083461

AN - SCOPUS:77954762644

SN - 1083-4419

VL - 40

SP - 1170

EP - 1175

JO - IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

JF - IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

IS - 4

ER -

L1-norm-based 2DPCA

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this