Discriminant feature extraction for parametric and non-parametric classifier

Chulhee Lee; David A. Landgrebe

doi:10.1109/ICSMC.1992.271598

Discriminant feature extraction for parametric and non-parametric classifier

Chulhee Lee, David A. Landgrebe

Department of Electrical and Electronic Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

In this paper, feature extraction is considered as preserving the value of the discriminant function for a given classifier which uses a posteriori probabilities P(ω|X) while reducing dimensionality. For classification minimizing Bayes' error, a posteriori probabilities would be the best features. In this feature space, the probability of error is the same as in the original space assuming Bayes' classifier. We consider feature extraction as eliminating features which have no impact on the value of the discriminant function and propose a feature extraction algorithm which eliminates those irrelevant features and retains only useful features. The proposed feature extraction algorithm does not deteriorate even when there is no difference in the mean vectors or no differences in the covariance matrices, and can be used for both parametric classifiers and non-parametric classifiers.

Original language	English
Title of host publication	1992 IEEE International Conference on Systems, Man, and Cybernetics
Subtitle of host publication	Emergent Innovations in Information Transfer Processing and Decision Making, SMC 1992
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1345-1350
Number of pages	6
ISBN (Electronic)	0780307208, 9780780307209
DOIs	https://doi.org/10.1109/ICSMC.1992.271598
Publication status	Published - 1992
Event	IEEE International Conference on Systems, Man, and Cybernetics, SMC 1992 - Chicago, United States Duration: 1992 Oct 18 → 1992 Oct 21

Publication series

Name	Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics
Volume	1992-January
ISSN (Print)	1062-922X

Other

Other	IEEE International Conference on Systems, Man, and Cybernetics, SMC 1992
Country/Territory	United States
City	Chicago
Period	92/10/18 → 92/10/21

Bibliographical note

Publisher Copyright:
© 1992 IEEE.

All Science Journal Classification (ASJC) codes

Electrical and Electronic Engineering
Control and Systems Engineering
Human-Computer Interaction

Access to Document

10.1109/ICSMC.1992.271598

Cite this

Lee, C., & Landgrebe, D. A. (1992). Discriminant feature extraction for parametric and non-parametric classifier. In 1992 IEEE International Conference on Systems, Man, and Cybernetics: Emergent Innovations in Information Transfer Processing and Decision Making, SMC 1992 (pp. 1345-1350). Article 271598 (Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics; Vol. 1992-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICSMC.1992.271598

Lee, Chulhee ; Landgrebe, David A. / Discriminant feature extraction for parametric and non-parametric classifier. 1992 IEEE International Conference on Systems, Man, and Cybernetics: Emergent Innovations in Information Transfer Processing and Decision Making, SMC 1992. Institute of Electrical and Electronics Engineers Inc., 1992. pp. 1345-1350 (Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics).

@inproceedings{f7337f5a059d4644861e69555e03aa18,

title = "Discriminant feature extraction for parametric and non-parametric classifier",

abstract = "In this paper, feature extraction is considered as preserving the value of the discriminant function for a given classifier which uses a posteriori probabilities P(ω|X) while reducing dimensionality. For classification minimizing Bayes' error, a posteriori probabilities would be the best features. In this feature space, the probability of error is the same as in the original space assuming Bayes' classifier. We consider feature extraction as eliminating features which have no impact on the value of the discriminant function and propose a feature extraction algorithm which eliminates those irrelevant features and retains only useful features. The proposed feature extraction algorithm does not deteriorate even when there is no difference in the mean vectors or no differences in the covariance matrices, and can be used for both parametric classifiers and non-parametric classifiers.",

author = "Chulhee Lee and Landgrebe, {David A.}",

note = "Publisher Copyright: {\textcopyright} 1992 IEEE.; IEEE International Conference on Systems, Man, and Cybernetics, SMC 1992 ; Conference date: 18-10-1992 Through 21-10-1992",

year = "1992",

doi = "10.1109/ICSMC.1992.271598",

language = "English",

series = "Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "1345--1350",

booktitle = "1992 IEEE International Conference on Systems, Man, and Cybernetics",

address = "United States",

}

Lee, C & Landgrebe, DA 1992, Discriminant feature extraction for parametric and non-parametric classifier. in 1992 IEEE International Conference on Systems, Man, and Cybernetics: Emergent Innovations in Information Transfer Processing and Decision Making, SMC 1992., 271598, Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics, vol. 1992-January, Institute of Electrical and Electronics Engineers Inc., pp. 1345-1350, IEEE International Conference on Systems, Man, and Cybernetics, SMC 1992, Chicago, United States, 92/10/18. https://doi.org/10.1109/ICSMC.1992.271598

Discriminant feature extraction for parametric and non-parametric classifier. / Lee, Chulhee; Landgrebe, David A.
1992 IEEE International Conference on Systems, Man, and Cybernetics: Emergent Innovations in Information Transfer Processing and Decision Making, SMC 1992. Institute of Electrical and Electronics Engineers Inc., 1992. p. 1345-1350 271598 (Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics; Vol. 1992-January).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Discriminant feature extraction for parametric and non-parametric classifier

AU - Lee, Chulhee

AU - Landgrebe, David A.

PY - 1992

Y1 - 1992

N2 - In this paper, feature extraction is considered as preserving the value of the discriminant function for a given classifier which uses a posteriori probabilities P(ω|X) while reducing dimensionality. For classification minimizing Bayes' error, a posteriori probabilities would be the best features. In this feature space, the probability of error is the same as in the original space assuming Bayes' classifier. We consider feature extraction as eliminating features which have no impact on the value of the discriminant function and propose a feature extraction algorithm which eliminates those irrelevant features and retains only useful features. The proposed feature extraction algorithm does not deteriorate even when there is no difference in the mean vectors or no differences in the covariance matrices, and can be used for both parametric classifiers and non-parametric classifiers.

AB - In this paper, feature extraction is considered as preserving the value of the discriminant function for a given classifier which uses a posteriori probabilities P(ω|X) while reducing dimensionality. For classification minimizing Bayes' error, a posteriori probabilities would be the best features. In this feature space, the probability of error is the same as in the original space assuming Bayes' classifier. We consider feature extraction as eliminating features which have no impact on the value of the discriminant function and propose a feature extraction algorithm which eliminates those irrelevant features and retains only useful features. The proposed feature extraction algorithm does not deteriorate even when there is no difference in the mean vectors or no differences in the covariance matrices, and can be used for both parametric classifiers and non-parametric classifiers.

UR - http://www.scopus.com/inward/record.url?scp=84960344808&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84960344808&partnerID=8YFLogxK

U2 - 10.1109/ICSMC.1992.271598

DO - 10.1109/ICSMC.1992.271598

M3 - Conference contribution

AN - SCOPUS:84960344808

T3 - Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics

SP - 1345

EP - 1350

BT - 1992 IEEE International Conference on Systems, Man, and Cybernetics

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - IEEE International Conference on Systems, Man, and Cybernetics, SMC 1992

Y2 - 18 October 1992 through 21 October 1992

ER -

Lee C, Landgrebe DA. Discriminant feature extraction for parametric and non-parametric classifier. In 1992 IEEE International Conference on Systems, Man, and Cybernetics: Emergent Innovations in Information Transfer Processing and Decision Making, SMC 1992. Institute of Electrical and Electronics Engineers Inc. 1992. p. 1345-1350. 271598. (Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics). doi: 10.1109/ICSMC.1992.271598

Discriminant feature extraction for parametric and non-parametric classifier

Abstract

Publication series

Other

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this