Kernel parameter optimization for kernel-based LDA methods

Jian Huang; Xiaoming Chen; Pong Chi YUEN; Jun Zhang; W. S. Chen; J. H. Lai

doi:10.1109/IJCNN.2008.4634350

Kernel parameter optimization for kernel-based LDA methods

Jian Huang^*, Xiaoming Chen, Pong Chi YUEN, Jun Zhang, W. S. Chen, J. H. Lai

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

2 Citations (Scopus)

Abstract

Kernel approach has been employed to solve classification problem with complex distribution by mapping the input space to higher dimensional feature space. However, one of the crucial factors in the Kernel approach is the choosing of kernel parameters which highly affect the performance and stability of the kernel-based learning methods. In view of this limitation, this paper adopts the Eigenvalue Stability Bounded Margin Maximization (ESBMM) algorithm to automatically tune the multiple kernel parameters for Kernel-based LDA methods. To demonstrate its effectiveness, the ESBMM algorithm has been extended and applied on two existing kernelbased LDA methods. Experimental results show that after applying the ESBMM algorithm, the performance of these two methods are both improved.

Original language	English
Title of host publication	2008 International Joint Conference on Neural Networks, IJCNN 2008
Pages	3840-3846
Number of pages	7
DOIs	https://doi.org/10.1109/IJCNN.2008.4634350
Publication status	Published - 2008
Event	2008 International Joint Conference on Neural Networks, IJCNN 2008 - Hong Kong, China Duration: 1 Jun 2008 → 8 Jun 2008

Publication series

Name	Proceedings of the International Joint Conference on Neural Networks

Conference

Conference	2008 International Joint Conference on Neural Networks, IJCNN 2008
Country/Territory	China
City	Hong Kong
Period	1/06/08 → 8/06/08

User-Defined Keywords

Face recognition
Kernel fisher discriminant
Kernel parameter
Stability

Access to Document

10.1109/IJCNN.2008.4634350

Cite this

@inproceedings{de23fdf2bfc542568a33f4e32ca23f54,

title = "Kernel parameter optimization for kernel-based LDA methods",

abstract = "Kernel approach has been employed to solve classification problem with complex distribution by mapping the input space to higher dimensional feature space. However, one of the crucial factors in the Kernel approach is the choosing of kernel parameters which highly affect the performance and stability of the kernel-based learning methods. In view of this limitation, this paper adopts the Eigenvalue Stability Bounded Margin Maximization (ESBMM) algorithm to automatically tune the multiple kernel parameters for Kernel-based LDA methods. To demonstrate its effectiveness, the ESBMM algorithm has been extended and applied on two existing kernelbased LDA methods. Experimental results show that after applying the ESBMM algorithm, the performance of these two methods are both improved.",

keywords = "Face recognition, Kernel fisher discriminant, Kernel parameter, Stability",

author = "Jian Huang and Xiaoming Chen and YUEN, {Pong Chi} and Jun Zhang and Chen, {W. S.} and Lai, {J. H.}",

year = "2008",

doi = "10.1109/IJCNN.2008.4634350",

language = "English",

isbn = "9781424418213",

series = "Proceedings of the International Joint Conference on Neural Networks",

pages = "3840--3846",

booktitle = "2008 International Joint Conference on Neural Networks, IJCNN 2008",

}

Huang, J, Chen, X, YUEN, PC, Zhang, J, Chen, WS & Lai, JH 2008, Kernel parameter optimization for kernel-based LDA methods. in 2008 International Joint Conference on Neural Networks, IJCNN 2008., 4634350, Proceedings of the International Joint Conference on Neural Networks, pp. 3840-3846, 2008 International Joint Conference on Neural Networks, IJCNN 2008, Hong Kong, China, 1/06/08. https://doi.org/10.1109/IJCNN.2008.4634350

Kernel parameter optimization for kernel-based LDA methods. / Huang, Jian; Chen, Xiaoming; YUEN, Pong Chi et al.
2008 International Joint Conference on Neural Networks, IJCNN 2008. 2008. p. 3840-3846 4634350 (Proceedings of the International Joint Conference on Neural Networks).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Kernel parameter optimization for kernel-based LDA methods

AU - Huang, Jian

AU - Chen, Xiaoming

AU - YUEN, Pong Chi

AU - Zhang, Jun

AU - Chen, W. S.

AU - Lai, J. H.

PY - 2008

Y1 - 2008

N2 - Kernel approach has been employed to solve classification problem with complex distribution by mapping the input space to higher dimensional feature space. However, one of the crucial factors in the Kernel approach is the choosing of kernel parameters which highly affect the performance and stability of the kernel-based learning methods. In view of this limitation, this paper adopts the Eigenvalue Stability Bounded Margin Maximization (ESBMM) algorithm to automatically tune the multiple kernel parameters for Kernel-based LDA methods. To demonstrate its effectiveness, the ESBMM algorithm has been extended and applied on two existing kernelbased LDA methods. Experimental results show that after applying the ESBMM algorithm, the performance of these two methods are both improved.

AB - Kernel approach has been employed to solve classification problem with complex distribution by mapping the input space to higher dimensional feature space. However, one of the crucial factors in the Kernel approach is the choosing of kernel parameters which highly affect the performance and stability of the kernel-based learning methods. In view of this limitation, this paper adopts the Eigenvalue Stability Bounded Margin Maximization (ESBMM) algorithm to automatically tune the multiple kernel parameters for Kernel-based LDA methods. To demonstrate its effectiveness, the ESBMM algorithm has been extended and applied on two existing kernelbased LDA methods. Experimental results show that after applying the ESBMM algorithm, the performance of these two methods are both improved.

KW - Face recognition

KW - Kernel fisher discriminant

KW - Kernel parameter

KW - Stability

UR - http://www.scopus.com/inward/record.url?scp=56349111496&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2008.4634350

DO - 10.1109/IJCNN.2008.4634350

M3 - Conference proceeding

AN - SCOPUS:56349111496

SN - 9781424418213

T3 - Proceedings of the International Joint Conference on Neural Networks

SP - 3840

EP - 3846

BT - 2008 International Joint Conference on Neural Networks, IJCNN 2008

T2 - 2008 International Joint Conference on Neural Networks, IJCNN 2008

Y2 - 1 June 2008 through 8 June 2008

ER -

Kernel parameter optimization for kernel-based LDA methods

Abstract

Publication series

Conference

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this