Dimensionality reduction using kernel pooled local discriminant information

Peng Zhang, Jing Peng, Carlotta Domeniconi

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

We study the use of kernel subspace methods for learning low-dimensional representations for classification. We propose a kernel pooled local discriminant subspace method and compare it against several competing techniques: generalized Fisher discriminant analysis (GDA) and kernel principal components analysis (KPCA) in classification problems. We evaluate the classification performance of the nearest-neighbor rule with each subspace representation. The experimental results demonstrate the efficacy of the kernel pooled local subspace method and the potential for substantial improvements over competing methods such as KPCA in some classification problems.

Original languageEnglish
Title of host publicationProceedings - 3rd IEEE International Conference on Data Mining, ICDM 2003
Pages701-704
Number of pages4
StatePublished - 1 Dec 2003
Event3rd IEEE International Conference on Data Mining, ICDM '03 - Melbourne, FL, United States
Duration: 19 Nov 200322 Nov 2003

Other

Other3rd IEEE International Conference on Data Mining, ICDM '03
CountryUnited States
CityMelbourne, FL
Period19/11/0322/11/03

Fingerprint

Principal component analysis
Discriminant analysis

Cite this

Zhang, P., Peng, J., & Domeniconi, C. (2003). Dimensionality reduction using kernel pooled local discriminant information. In Proceedings - 3rd IEEE International Conference on Data Mining, ICDM 2003 (pp. 701-704)
Zhang, Peng ; Peng, Jing ; Domeniconi, Carlotta. / Dimensionality reduction using kernel pooled local discriminant information. Proceedings - 3rd IEEE International Conference on Data Mining, ICDM 2003. 2003. pp. 701-704
@inproceedings{94122c56b75546b4bb8bdad162dc0ae4,
title = "Dimensionality reduction using kernel pooled local discriminant information",
abstract = "We study the use of kernel subspace methods for learning low-dimensional representations for classification. We propose a kernel pooled local discriminant subspace method and compare it against several competing techniques: generalized Fisher discriminant analysis (GDA) and kernel principal components analysis (KPCA) in classification problems. We evaluate the classification performance of the nearest-neighbor rule with each subspace representation. The experimental results demonstrate the efficacy of the kernel pooled local subspace method and the potential for substantial improvements over competing methods such as KPCA in some classification problems.",
author = "Peng Zhang and Jing Peng and Carlotta Domeniconi",
year = "2003",
month = "12",
day = "1",
language = "English",
isbn = "0769519784",
pages = "701--704",
booktitle = "Proceedings - 3rd IEEE International Conference on Data Mining, ICDM 2003",

}

Zhang, P, Peng, J & Domeniconi, C 2003, Dimensionality reduction using kernel pooled local discriminant information. in Proceedings - 3rd IEEE International Conference on Data Mining, ICDM 2003. pp. 701-704, 3rd IEEE International Conference on Data Mining, ICDM '03, Melbourne, FL, United States, 19/11/03.

Dimensionality reduction using kernel pooled local discriminant information. / Zhang, Peng; Peng, Jing; Domeniconi, Carlotta.

Proceedings - 3rd IEEE International Conference on Data Mining, ICDM 2003. 2003. p. 701-704.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

TY - GEN

T1 - Dimensionality reduction using kernel pooled local discriminant information

AU - Zhang, Peng

AU - Peng, Jing

AU - Domeniconi, Carlotta

PY - 2003/12/1

Y1 - 2003/12/1

N2 - We study the use of kernel subspace methods for learning low-dimensional representations for classification. We propose a kernel pooled local discriminant subspace method and compare it against several competing techniques: generalized Fisher discriminant analysis (GDA) and kernel principal components analysis (KPCA) in classification problems. We evaluate the classification performance of the nearest-neighbor rule with each subspace representation. The experimental results demonstrate the efficacy of the kernel pooled local subspace method and the potential for substantial improvements over competing methods such as KPCA in some classification problems.

AB - We study the use of kernel subspace methods for learning low-dimensional representations for classification. We propose a kernel pooled local discriminant subspace method and compare it against several competing techniques: generalized Fisher discriminant analysis (GDA) and kernel principal components analysis (KPCA) in classification problems. We evaluate the classification performance of the nearest-neighbor rule with each subspace representation. The experimental results demonstrate the efficacy of the kernel pooled local subspace method and the potential for substantial improvements over competing methods such as KPCA in some classification problems.

UR - http://www.scopus.com/inward/record.url?scp=78149306861&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0769519784

SN - 9780769519784

SP - 701

EP - 704

BT - Proceedings - 3rd IEEE International Conference on Data Mining, ICDM 2003

ER -

Zhang P, Peng J, Domeniconi C. Dimensionality reduction using kernel pooled local discriminant information. In Proceedings - 3rd IEEE International Conference on Data Mining, ICDM 2003. 2003. p. 701-704