Essence of kernel Fisher discriminant: KPCA plus LDA

Jian Yang; Zhong Jin; Jing Yu Yang; David Zhang; Alejandro F. Frangi

doi:10.1016/j.patcog.2003.10.015

Essence of kernel Fisher discriminant: KPCA plus LDA

Jian Yang, Zhong Jin, Jing Yu Yang, David Zhang, Alejandro F. Frangi

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

157 Citations (Scopus)

Abstract

In this paper, the method of kernel Fisher discriminant (KFD) is analyzed and its nature is revealed, i.e., KFD is equivalent to kernel principal component analysis (KPCA) plus Fisher linear discriminant analysis (LDA). Based on this result, a more transparent KFD algorithm is proposed. That is, KPCA is first performed and then LDA is used for a second feature extraction in the KPCA-transformed space. Finally, the effectiveness of the proposed algorithm is verified using the CENPARMI handwritten numeral database. © 2003 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.

Original language	English
Pages (from-to)	2097-2100
Journal	Pattern Recognition
Volume	37
Issue number	10
DOIs	https://doi.org/10.1016/j.patcog.2003.10.015
Publication status	Published - 1 Oct 2004

Keywords

Feature extraction
Fisher linear discriminant analysis
Handwritten numeral recognition
Kernel-based methods
Principal component analysis

Access to Document

10.1016/j.patcog.2003.10.015

Fingerprint Dive into the research topics of 'Essence of kernel Fisher discriminant: KPCA plus LDA'. Together they form a unique fingerprint.

Cite this

@article{0b6c6e695f3e403e97f8ade907596883,

title = "Essence of kernel Fisher discriminant: KPCA plus LDA",

abstract = "In this paper, the method of kernel Fisher discriminant (KFD) is analyzed and its nature is revealed, i.e., KFD is equivalent to kernel principal component analysis (KPCA) plus Fisher linear discriminant analysis (LDA). Based on this result, a more transparent KFD algorithm is proposed. That is, KPCA is first performed and then LDA is used for a second feature extraction in the KPCA-transformed space. Finally, the effectiveness of the proposed algorithm is verified using the CENPARMI handwritten numeral database. {\textcopyright} 2003 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.",

keywords = "Feature extraction, Fisher linear discriminant analysis, Handwritten numeral recognition, Kernel-based methods, Principal component analysis",

author = "Jian Yang and Zhong Jin and Yang, {Jing Yu} and David Zhang and Frangi, {Alejandro F.}",

year = "2004",

month = oct,

day = "1",

doi = "10.1016/j.patcog.2003.10.015",

language = "English",

volume = "37",

pages = "2097--2100",

journal = "Pattern Recognition",

issn = "0031-3203",

publisher = "Elsevier Limited",

number = "10",

}

TY - JOUR

T1 - Essence of kernel Fisher discriminant: KPCA plus LDA

AU - Yang, Jian

AU - Jin, Zhong

AU - Yang, Jing Yu

AU - Zhang, David

AU - Frangi, Alejandro F.

PY - 2004/10/1

Y1 - 2004/10/1

N2 - In this paper, the method of kernel Fisher discriminant (KFD) is analyzed and its nature is revealed, i.e., KFD is equivalent to kernel principal component analysis (KPCA) plus Fisher linear discriminant analysis (LDA). Based on this result, a more transparent KFD algorithm is proposed. That is, KPCA is first performed and then LDA is used for a second feature extraction in the KPCA-transformed space. Finally, the effectiveness of the proposed algorithm is verified using the CENPARMI handwritten numeral database. © 2003 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.

AB - In this paper, the method of kernel Fisher discriminant (KFD) is analyzed and its nature is revealed, i.e., KFD is equivalent to kernel principal component analysis (KPCA) plus Fisher linear discriminant analysis (LDA). Based on this result, a more transparent KFD algorithm is proposed. That is, KPCA is first performed and then LDA is used for a second feature extraction in the KPCA-transformed space. Finally, the effectiveness of the proposed algorithm is verified using the CENPARMI handwritten numeral database. © 2003 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.

KW - Feature extraction

KW - Fisher linear discriminant analysis

KW - Handwritten numeral recognition

KW - Kernel-based methods

KW - Principal component analysis

U2 - 10.1016/j.patcog.2003.10.015

DO - 10.1016/j.patcog.2003.10.015

M3 - Article

VL - 37

SP - 2097

EP - 2100

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

IS - 10

ER -