Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information

Yoshifusa Ito; Cidambi Srinivasan

doi:10.1007/3-540-44668-0_19

Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information

Yoshifusa Ito, Cidambi Srinivasan

Dr. Bing Zhang Department of Statistics

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

2 Scopus citations

Abstract

Following general arguments on approximation Bayesian discriminant functions by neural networks, rigorously proved is that a three layered neural network, having rather a small number of hidden layer units, can approximate the Bayesian discriminant function for the two category classification if the log ratio of the a posteriori probability is a polynomial. The accuracy of approximation is measured by the Kullback- Leibler information. An extension to the multi-category case is also discussed.

Original language	English
Title of host publication	Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings
Editors	Kurt Hornik, Georg Dorffner, Horst Bischof
Pages	135-140
Number of pages	6
DOIs	https://doi.org/10.1007/3-540-44668-0_19
State	Published - 2001
Event	International Conference on Artificial Neural Networks, ICANN 2001 - Vienna, Austria Duration: Aug 21 2001 → Aug 25 2001

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	2130
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Conference on Artificial Neural Networks, ICANN 2001
Country/Territory	Austria
City	Vienna
Period	8/21/01 → 8/25/01

Bibliographical note

Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2001.

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/3-540-44668-0_19

Cite this

Ito, Y., & Srinivasan, C. (2001). Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information. In K. Hornik, G. Dorffner, & H. Bischof (Eds.), Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings (pp. 135-140). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2130). https://doi.org/10.1007/3-540-44668-0_19

Ito, Yoshifusa ; Srinivasan, Cidambi. / Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information. Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings. editor / Kurt Hornik ; Georg Dorffner ; Horst Bischof. 2001. pp. 135-140 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{2620257e9b5a464d97c37465887fa93d,

title = "Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information",

abstract = "Following general arguments on approximation Bayesian discriminant functions by neural networks, rigorously proved is that a three layered neural network, having rather a small number of hidden layer units, can approximate the Bayesian discriminant function for the two category classification if the log ratio of the a posteriori probability is a polynomial. The accuracy of approximation is measured by the Kullback- Leibler information. An extension to the multi-category case is also discussed.",

author = "Yoshifusa Ito and Cidambi Srinivasan",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 2001.; International Conference on Artificial Neural Networks, ICANN 2001 ; Conference date: 21-08-2001 Through 25-08-2001",

year = "2001",

doi = "10.1007/3-540-44668-0_19",

language = "English",

isbn = "3540424865",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "135--140",

editor = "Kurt Hornik and Georg Dorffner and Horst Bischof",

booktitle = "Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings",

}

Ito, Y & Srinivasan, C 2001, Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information. in K Hornik, G Dorffner & H Bischof (eds), Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 2130, pp. 135-140, International Conference on Artificial Neural Networks, ICANN 2001, Vienna, Austria, 8/21/01. https://doi.org/10.1007/3-540-44668-0_19

Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information. / Ito, Yoshifusa; Srinivasan, Cidambi.
Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings. ed. / Kurt Hornik; Georg Dorffner; Horst Bischof. 2001. p. 135-140 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2130).

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

TY - GEN

T1 - Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information

AU - Ito, Yoshifusa

AU - Srinivasan, Cidambi

N1 - Publisher Copyright: © Springer-Verlag Berlin Heidelberg 2001.

PY - 2001

Y1 - 2001

N2 - Following general arguments on approximation Bayesian discriminant functions by neural networks, rigorously proved is that a three layered neural network, having rather a small number of hidden layer units, can approximate the Bayesian discriminant function for the two category classification if the log ratio of the a posteriori probability is a polynomial. The accuracy of approximation is measured by the Kullback- Leibler information. An extension to the multi-category case is also discussed.

AB - Following general arguments on approximation Bayesian discriminant functions by neural networks, rigorously proved is that a three layered neural network, having rather a small number of hidden layer units, can approximate the Bayesian discriminant function for the two category classification if the log ratio of the a posteriori probability is a polynomial. The accuracy of approximation is measured by the Kullback- Leibler information. An extension to the multi-category case is also discussed.

UR - http://www.scopus.com/inward/record.url?scp=23044525372&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=23044525372&partnerID=8YFLogxK

U2 - 10.1007/3-540-44668-0_19

DO - 10.1007/3-540-44668-0_19

M3 - Conference contribution

AN - SCOPUS:23044525372

SN - 3540424865

SN - 9783540446682

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 135

EP - 140

BT - Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings

A2 - Hornik, Kurt

A2 - Dorffner, Georg

A2 - Bischof, Horst

T2 - International Conference on Artificial Neural Networks, ICANN 2001

Y2 - 21 August 2001 through 25 August 2001

ER -

Ito Y, Srinivasan C. Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information. In Hornik K, Dorffner G, Bischof H, editors, Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings. 2001. p. 135-140. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/3-540-44668-0_19

Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information

Abstract

Publication series

Conference

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this