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

Producción científica: Conference contribution › revisión exhaustiva

2 Citas (Scopus)

Resumen

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.

Idioma original	English
Título de la publicación alojada	Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings
Editores	Kurt Hornik, Georg Dorffner, Horst Bischof
Páginas	135-140
Número de páginas	6
DOI	https://doi.org/10.1007/3-540-44668-0_19
Estado	Published - 2001
Evento	International Conference on Artificial Neural Networks, ICANN 2001 - Vienna, Austria Duración: ago 21 2001 → ago 25 2001

Serie de la publicación

Nombre	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen	2130
ISSN (versión impresa)	0302-9743
ISSN (versión digital)	1611-3349

Conference

Conference	International Conference on Artificial Neural Networks, ICANN 2001
País/Territorio	Austria
Ciudad	Vienna
Período	8/21/01 → 8/25/01

Nota bibliográfica

Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2001.

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Acceder al documento

10.1007/3-540-44668-0_19

Otros archivos y enlaces

Citar esto

Ito, Y., & Srinivasan, C. (2001). Approximation of Bayesian discriminant function by neural networks in terms of Kullback-Leibler information. En 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. En 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).

Producción científica: Conference contribution › revisión exhaustiva

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 - https://www.scopus.com/pages/publications/23044525372

UR - https://www.scopus.com/pages/publications/23044525372#tab=citedBy

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. En Hornik K, Dorffner G, Bischof H, editores, 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

Resumen

Serie de la publicación

Conference

Nota bibliográfica

ASJC Scopus subject areas

Acceder al documento

Otros archivos y enlaces

Huella

Citar esto