Fast statistical learning with Kernel-based simple-fda

K. Nakaura, S. Karungaru, T. Akashi, Yasue Mitsukura, M. Fukumi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, new statistical learning algorithms with kernel function are presented. Recently, iterative learning algorithms for obtaining eigenvectors in the principal component analysis (PCA) have been presented in the field of pattern recognition and neural network. However, the Fisher linear discriminant analysis (FLDA) has been used in many fields, especially face image analysis. The drawback of FLDA is a long computational time based on a large-sized covariance matrix and the issue that the within-class covariance matrix is usually singular. In order to overcome this difficulty, we proposed the feature generation method Simple-FLDA which is approximately derived from geometrical interpretation of FLDA. This algorithm is similar to Simple-PCA and does not need matrix operation. In this paper, new statistical kernel based learning algorithms are presented. They are extended versions of Simple-PCA and Simple-FLDA to nonlinear space using the kernel function. Their preliminary simulation results are given for a simple face recognition problem.

Original languageEnglish
Title of host publicationSITIS 2008 - Proceedings of the 4th International Conference on Signal Image Technology and Internet Based Systems
Pages333-337
Number of pages5
DOIs
Publication statusPublished - 2008
Externally publishedYes
Event4th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2008 - Bali, Indonesia
Duration: 2008 Nov 302008 Dec 3

Other

Other4th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2008
CountryIndonesia
CityBali
Period08/11/3008/12/3

Fingerprint

Discriminant analysis
Principal component analysis
Learning algorithms
Covariance matrix
Face recognition
Eigenvalues and eigenfunctions
Image analysis
Pattern recognition
Neural networks

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Software

Cite this

Nakaura, K., Karungaru, S., Akashi, T., Mitsukura, Y., & Fukumi, M. (2008). Fast statistical learning with Kernel-based simple-fda. In SITIS 2008 - Proceedings of the 4th International Conference on Signal Image Technology and Internet Based Systems (pp. 333-337). [4725823] https://doi.org/10.1109/SITIS.2008.52

Fast statistical learning with Kernel-based simple-fda. / Nakaura, K.; Karungaru, S.; Akashi, T.; Mitsukura, Yasue; Fukumi, M.

SITIS 2008 - Proceedings of the 4th International Conference on Signal Image Technology and Internet Based Systems. 2008. p. 333-337 4725823.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nakaura, K, Karungaru, S, Akashi, T, Mitsukura, Y & Fukumi, M 2008, Fast statistical learning with Kernel-based simple-fda. in SITIS 2008 - Proceedings of the 4th International Conference on Signal Image Technology and Internet Based Systems., 4725823, pp. 333-337, 4th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2008, Bali, Indonesia, 08/11/30. https://doi.org/10.1109/SITIS.2008.52
Nakaura K, Karungaru S, Akashi T, Mitsukura Y, Fukumi M. Fast statistical learning with Kernel-based simple-fda. In SITIS 2008 - Proceedings of the 4th International Conference on Signal Image Technology and Internet Based Systems. 2008. p. 333-337. 4725823 https://doi.org/10.1109/SITIS.2008.52
Nakaura, K. ; Karungaru, S. ; Akashi, T. ; Mitsukura, Yasue ; Fukumi, M. / Fast statistical learning with Kernel-based simple-fda. SITIS 2008 - Proceedings of the 4th International Conference on Signal Image Technology and Internet Based Systems. 2008. pp. 333-337
@inproceedings{7f4dd5fe328a45ff8aafe4c45ea8728b,
title = "Fast statistical learning with Kernel-based simple-fda",
abstract = "In this paper, new statistical learning algorithms with kernel function are presented. Recently, iterative learning algorithms for obtaining eigenvectors in the principal component analysis (PCA) have been presented in the field of pattern recognition and neural network. However, the Fisher linear discriminant analysis (FLDA) has been used in many fields, especially face image analysis. The drawback of FLDA is a long computational time based on a large-sized covariance matrix and the issue that the within-class covariance matrix is usually singular. In order to overcome this difficulty, we proposed the feature generation method Simple-FLDA which is approximately derived from geometrical interpretation of FLDA. This algorithm is similar to Simple-PCA and does not need matrix operation. In this paper, new statistical kernel based learning algorithms are presented. They are extended versions of Simple-PCA and Simple-FLDA to nonlinear space using the kernel function. Their preliminary simulation results are given for a simple face recognition problem.",
author = "K. Nakaura and S. Karungaru and T. Akashi and Yasue Mitsukura and M. Fukumi",
year = "2008",
doi = "10.1109/SITIS.2008.52",
language = "English",
isbn = "9780769534930",
pages = "333--337",
booktitle = "SITIS 2008 - Proceedings of the 4th International Conference on Signal Image Technology and Internet Based Systems",

}

TY - GEN

T1 - Fast statistical learning with Kernel-based simple-fda

AU - Nakaura, K.

AU - Karungaru, S.

AU - Akashi, T.

AU - Mitsukura, Yasue

AU - Fukumi, M.

PY - 2008

Y1 - 2008

N2 - In this paper, new statistical learning algorithms with kernel function are presented. Recently, iterative learning algorithms for obtaining eigenvectors in the principal component analysis (PCA) have been presented in the field of pattern recognition and neural network. However, the Fisher linear discriminant analysis (FLDA) has been used in many fields, especially face image analysis. The drawback of FLDA is a long computational time based on a large-sized covariance matrix and the issue that the within-class covariance matrix is usually singular. In order to overcome this difficulty, we proposed the feature generation method Simple-FLDA which is approximately derived from geometrical interpretation of FLDA. This algorithm is similar to Simple-PCA and does not need matrix operation. In this paper, new statistical kernel based learning algorithms are presented. They are extended versions of Simple-PCA and Simple-FLDA to nonlinear space using the kernel function. Their preliminary simulation results are given for a simple face recognition problem.

AB - In this paper, new statistical learning algorithms with kernel function are presented. Recently, iterative learning algorithms for obtaining eigenvectors in the principal component analysis (PCA) have been presented in the field of pattern recognition and neural network. However, the Fisher linear discriminant analysis (FLDA) has been used in many fields, especially face image analysis. The drawback of FLDA is a long computational time based on a large-sized covariance matrix and the issue that the within-class covariance matrix is usually singular. In order to overcome this difficulty, we proposed the feature generation method Simple-FLDA which is approximately derived from geometrical interpretation of FLDA. This algorithm is similar to Simple-PCA and does not need matrix operation. In this paper, new statistical kernel based learning algorithms are presented. They are extended versions of Simple-PCA and Simple-FLDA to nonlinear space using the kernel function. Their preliminary simulation results are given for a simple face recognition problem.

UR - http://www.scopus.com/inward/record.url?scp=60349108293&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=60349108293&partnerID=8YFLogxK

U2 - 10.1109/SITIS.2008.52

DO - 10.1109/SITIS.2008.52

M3 - Conference contribution

AN - SCOPUS:60349108293

SN - 9780769534930

SP - 333

EP - 337

BT - SITIS 2008 - Proceedings of the 4th International Conference on Signal Image Technology and Internet Based Systems

ER -