Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes

Yoshifusa Ito; Hiroyuki Izumi; Cidambi Srinivasan

doi:10.1007/978-3-642-34500-5_4

Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes

Yoshifusa Ito, Hiroyuki Izumi, Cidambi Srinivasan

Dr. Bing Zhang Department of Statistics

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

Abstract

We construct a one-hidden-layer neural network capable of learning simultaneously several Bayesian discriminant functions and converting them to the corresponding Mahalanobis discriminant functions in the two-category, normal-distribution case. The Bayesian discriminant functions correspond to the respective situations on which the priors and means depend. The algorithm is characterized by the use of the inner potential of the output unit and additional several memory nodes. It is remarkably simpler when compared with our previous algorithm.

Original language	English
Title of host publication	Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings
Pages	25-33
Number of pages	9
Edition	PART 5
DOIs	https://doi.org/10.1007/978-3-642-34500-5_4
State	Published - 2012
Event	19th International Conference on Neural Information Processing, ICONIP 2012 - Doha, Qatar Duration: Nov 12 2012 → Nov 15 2012

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 5
Volume	7667 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	19th International Conference on Neural Information Processing, ICONIP 2012
Country/Territory	Qatar
City	Doha
Period	11/12/12 → 11/15/12

Keywords

Bayesian
Discriminant function
Mahalanobis
Simultaneous learning

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-642-34500-5_4

Cite this

Ito, Y., Izumi, H., & Srinivasan, C. (2012). Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes. In Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings (PART 5 ed., pp. 25-33). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7667 LNCS, No. PART 5). https://doi.org/10.1007/978-3-642-34500-5_4

Ito, Yoshifusa ; Izumi, Hiroyuki ; Srinivasan, Cidambi. / Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes. Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings. PART 5. ed. 2012. pp. 25-33 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 5).

@inproceedings{1a70a35bbb014b579f1c69c87c339b67,

title = "Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes",

abstract = "We construct a one-hidden-layer neural network capable of learning simultaneously several Bayesian discriminant functions and converting them to the corresponding Mahalanobis discriminant functions in the two-category, normal-distribution case. The Bayesian discriminant functions correspond to the respective situations on which the priors and means depend. The algorithm is characterized by the use of the inner potential of the output unit and additional several memory nodes. It is remarkably simpler when compared with our previous algorithm.",

keywords = "Bayesian, Discriminant function, Mahalanobis, Simultaneous learning",

author = "Yoshifusa Ito and Hiroyuki Izumi and Cidambi Srinivasan",

year = "2012",

doi = "10.1007/978-3-642-34500-5_4",

language = "English",

isbn = "9783642344992",

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

number = "PART 5",

pages = "25--33",

booktitle = "Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings",

edition = "PART 5",

note = "19th International Conference on Neural Information Processing, ICONIP 2012 ; Conference date: 12-11-2012 Through 15-11-2012",

}

Ito, Y, Izumi, H & Srinivasan, C 2012, Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes. in Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings. PART 5 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 5, vol. 7667 LNCS, pp. 25-33, 19th International Conference on Neural Information Processing, ICONIP 2012, Doha, Qatar, 11/12/12. https://doi.org/10.1007/978-3-642-34500-5_4

Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes. / Ito, Yoshifusa; Izumi, Hiroyuki; Srinivasan, Cidambi.
Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings. PART 5. ed. 2012. p. 25-33 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7667 LNCS, No. PART 5).

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

TY - GEN

T1 - Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes

AU - Ito, Yoshifusa

AU - Izumi, Hiroyuki

AU - Srinivasan, Cidambi

PY - 2012

Y1 - 2012

N2 - We construct a one-hidden-layer neural network capable of learning simultaneously several Bayesian discriminant functions and converting them to the corresponding Mahalanobis discriminant functions in the two-category, normal-distribution case. The Bayesian discriminant functions correspond to the respective situations on which the priors and means depend. The algorithm is characterized by the use of the inner potential of the output unit and additional several memory nodes. It is remarkably simpler when compared with our previous algorithm.

AB - We construct a one-hidden-layer neural network capable of learning simultaneously several Bayesian discriminant functions and converting them to the corresponding Mahalanobis discriminant functions in the two-category, normal-distribution case. The Bayesian discriminant functions correspond to the respective situations on which the priors and means depend. The algorithm is characterized by the use of the inner potential of the output unit and additional several memory nodes. It is remarkably simpler when compared with our previous algorithm.

KW - Bayesian

KW - Discriminant function

KW - Mahalanobis

KW - Simultaneous learning

UR - http://www.scopus.com/inward/record.url?scp=84869077891&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84869077891&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-34500-5_4

DO - 10.1007/978-3-642-34500-5_4

M3 - Conference contribution

AN - SCOPUS:84869077891

SN - 9783642344992

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

SP - 25

EP - 33

BT - Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings

T2 - 19th International Conference on Neural Information Processing, ICONIP 2012

Y2 - 12 November 2012 through 15 November 2012

ER -

Ito Y, Izumi H, Srinivasan C. Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes. In Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings. PART 5 ed. 2012. p. 25-33. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 5). doi: 10.1007/978-3-642-34500-5_4

Simultaneous learning of several Bayesian and mahalanobis discriminant functions by a neural network with memory nodes

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this