Combining classifiers based on minimization of a Bayes error rate

Hee Joong Kang; Seong Whan Lee

doi:10.1109/ICDAR.1999.791808

Combining classifiers based on minimization of a Bayes error rate

Hee Joong Kang, Seong Whan Lee

Department of Brain and Cognitive Engineering

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

23 Citations (Scopus)

Abstract

In order to raise a class discrimination power by combining multiple classifiers, the upper bound of a Bayes error rate bounded by the conditional entropy of a class variable and decision variables should be minimized. Wang and Wong (1979) proposed a tree dependence approximation scheme of a high order probability distribution composed of those variables, based on minimizing the upper bound. In addition to that, this paper presents an extended approximation scheme dealing with higher order dependency. Multiple classifiers recognizing unconstrained handwritten numerals were combined by the proposed approximation scheme based on the minimization of the Bayes error rate, and the high recognition rates were obtained by them.

Original language	English
Title of host publication	Proceedings of the 5th International Conference on Document Analysis and Recognition, ICDAR 1999
Publisher	IEEE Computer Society
Pages	398-401
Number of pages	4
ISBN (Electronic)	0769503187
DOIs	https://doi.org/10.1109/ICDAR.1999.791808
Publication status	Published - 1999
Event	5th International Conference on Document Analysis and Recognition, ICDAR 1999 - Bangalore, India Duration: 1999 Sept 20 → 1999 Sept 22

Publication series

Name	Proceedings of the International Conference on Document Analysis and Recognition, ICDAR
ISSN (Print)	1520-5363

Other

Other	5th International Conference on Document Analysis and Recognition, ICDAR 1999
Country/Territory	India
City	Bangalore
Period	99/9/20 → 99/9/22

ASJC Scopus subject areas

Computer Vision and Pattern Recognition

Access to Document

10.1109/ICDAR.1999.791808

Cite this

Kang, H. J., & Lee, S. W. (1999). Combining classifiers based on minimization of a Bayes error rate. In Proceedings of the 5th International Conference on Document Analysis and Recognition, ICDAR 1999 (pp. 398-401). [791808] (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR). IEEE Computer Society. https://doi.org/10.1109/ICDAR.1999.791808

Combining classifiers based on minimization of a Bayes error rate. / Kang, Hee Joong; Lee, Seong Whan.
Proceedings of the 5th International Conference on Document Analysis and Recognition, ICDAR 1999. IEEE Computer Society, 1999. p. 398-401 791808 (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR).

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

Kang, HJ & Lee, SW 1999, Combining classifiers based on minimization of a Bayes error rate. in Proceedings of the 5th International Conference on Document Analysis and Recognition, ICDAR 1999., 791808, Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, IEEE Computer Society, pp. 398-401, 5th International Conference on Document Analysis and Recognition, ICDAR 1999, Bangalore, India, 99/9/20. https://doi.org/10.1109/ICDAR.1999.791808

@inproceedings{b4734da3ca0b45e3a2685f3260e88ec0,

title = "Combining classifiers based on minimization of a Bayes error rate",

abstract = "In order to raise a class discrimination power by combining multiple classifiers, the upper bound of a Bayes error rate bounded by the conditional entropy of a class variable and decision variables should be minimized. Wang and Wong (1979) proposed a tree dependence approximation scheme of a high order probability distribution composed of those variables, based on minimizing the upper bound. In addition to that, this paper presents an extended approximation scheme dealing with higher order dependency. Multiple classifiers recognizing unconstrained handwritten numerals were combined by the proposed approximation scheme based on the minimization of the Bayes error rate, and the high recognition rates were obtained by them.",

author = "Kang, {Hee Joong} and Lee, {Seong Whan}",

note = "Publisher Copyright: {\textcopyright} 1999 IEEE.; 5th International Conference on Document Analysis and Recognition, ICDAR 1999 ; Conference date: 20-09-1999 Through 22-09-1999",

year = "1999",

doi = "10.1109/ICDAR.1999.791808",

language = "English",

series = "Proceedings of the International Conference on Document Analysis and Recognition, ICDAR",

publisher = "IEEE Computer Society",

pages = "398--401",

booktitle = "Proceedings of the 5th International Conference on Document Analysis and Recognition, ICDAR 1999",

}

TY - GEN

T1 - Combining classifiers based on minimization of a Bayes error rate

AU - Kang, Hee Joong

AU - Lee, Seong Whan

PY - 1999

Y1 - 1999

N2 - In order to raise a class discrimination power by combining multiple classifiers, the upper bound of a Bayes error rate bounded by the conditional entropy of a class variable and decision variables should be minimized. Wang and Wong (1979) proposed a tree dependence approximation scheme of a high order probability distribution composed of those variables, based on minimizing the upper bound. In addition to that, this paper presents an extended approximation scheme dealing with higher order dependency. Multiple classifiers recognizing unconstrained handwritten numerals were combined by the proposed approximation scheme based on the minimization of the Bayes error rate, and the high recognition rates were obtained by them.

AB - In order to raise a class discrimination power by combining multiple classifiers, the upper bound of a Bayes error rate bounded by the conditional entropy of a class variable and decision variables should be minimized. Wang and Wong (1979) proposed a tree dependence approximation scheme of a high order probability distribution composed of those variables, based on minimizing the upper bound. In addition to that, this paper presents an extended approximation scheme dealing with higher order dependency. Multiple classifiers recognizing unconstrained handwritten numerals were combined by the proposed approximation scheme based on the minimization of the Bayes error rate, and the high recognition rates were obtained by them.

UR - http://www.scopus.com/inward/record.url?scp=37949009754&partnerID=8YFLogxK

U2 - 10.1109/ICDAR.1999.791808

DO - 10.1109/ICDAR.1999.791808

M3 - Conference contribution

AN - SCOPUS:37949009754

T3 - Proceedings of the International Conference on Document Analysis and Recognition, ICDAR

SP - 398

EP - 401

BT - Proceedings of the 5th International Conference on Document Analysis and Recognition, ICDAR 1999

PB - IEEE Computer Society

T2 - 5th International Conference on Document Analysis and Recognition, ICDAR 1999

Y2 - 20 September 1999 through 22 September 1999

ER -

Combining classifiers based on minimization of a Bayes error rate

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this