Discriminant analysis by a neural network with mahalanobis distance

Yoshifusa Ito; Cidambi Srinivasan; Hiroyuki Izumi

doi:10.1007/11840930_36

Discriminant analysis by a neural network with mahalanobis distance

Yoshifusa Ito, Cidambi Srinivasan, Hiroyuki Izumi

Dr. Bing Zhang Department of Statistics

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

12 Scopus citations

Abstract

We propose a neural network which can approximate Mahalanobis discriminant functions after being trained. It can be realized if a Bayesian neural network is equipped with two additional subnetworks. The training is performed sequentially and, hence, the past teacher signals need not be memorized. In this paper, we treat the two-category normal-distribution case. The results of simple simulations are included.

Original language	English
Title of host publication	Artificial Neural Networks, ICANN 2006 - 16th International Conference, Proceedings
Pages	350-360
Number of pages	11
DOIs	https://doi.org/10.1007/11840930_36
State	Published - 2006
Event	16th International Conference on Artificial Neural Networks, ICANN 2006 - Athens, Greece Duration: Sep 10 2006 → Sep 14 2006

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	4132 LNCS - II
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th International Conference on Artificial Neural Networks, ICANN 2006
Country/Territory	Greece
City	Athens
Period	9/10/06 → 9/14/06

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/11840930_36

Cite this

Ito, Y., Srinivasan, C., & Izumi, H. (2006). Discriminant analysis by a neural network with mahalanobis distance. In Artificial Neural Networks, ICANN 2006 - 16th International Conference, Proceedings (pp. 350-360). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4132 LNCS - II). https://doi.org/10.1007/11840930_36

@inproceedings{2099794eb52a45f29e185e4c52afd999,

title = "Discriminant analysis by a neural network with mahalanobis distance",

abstract = "We propose a neural network which can approximate Mahalanobis discriminant functions after being trained. It can be realized if a Bayesian neural network is equipped with two additional subnetworks. The training is performed sequentially and, hence, the past teacher signals need not be memorized. In this paper, we treat the two-category normal-distribution case. The results of simple simulations are included.",

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

year = "2006",

doi = "10.1007/11840930_36",

language = "English",

isbn = "3540388710",

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

pages = "350--360",

booktitle = "Artificial Neural Networks, ICANN 2006 - 16th International Conference, Proceedings",

note = "16th International Conference on Artificial Neural Networks, ICANN 2006 ; Conference date: 10-09-2006 Through 14-09-2006",

}

Ito, Y, Srinivasan, C & Izumi, H 2006, Discriminant analysis by a neural network with mahalanobis distance. in Artificial Neural Networks, ICANN 2006 - 16th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4132 LNCS - II, pp. 350-360, 16th International Conference on Artificial Neural Networks, ICANN 2006, Athens, Greece, 9/10/06. https://doi.org/10.1007/11840930_36

Discriminant analysis by a neural network with mahalanobis distance. / Ito, Yoshifusa; Srinivasan, Cidambi; Izumi, Hiroyuki.
Artificial Neural Networks, ICANN 2006 - 16th International Conference, Proceedings. 2006. p. 350-360 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4132 LNCS - II).

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

TY - GEN

T1 - Discriminant analysis by a neural network with mahalanobis distance

AU - Ito, Yoshifusa

AU - Srinivasan, Cidambi

AU - Izumi, Hiroyuki

PY - 2006

Y1 - 2006

N2 - We propose a neural network which can approximate Mahalanobis discriminant functions after being trained. It can be realized if a Bayesian neural network is equipped with two additional subnetworks. The training is performed sequentially and, hence, the past teacher signals need not be memorized. In this paper, we treat the two-category normal-distribution case. The results of simple simulations are included.

AB - We propose a neural network which can approximate Mahalanobis discriminant functions after being trained. It can be realized if a Bayesian neural network is equipped with two additional subnetworks. The training is performed sequentially and, hence, the past teacher signals need not be memorized. In this paper, we treat the two-category normal-distribution case. The results of simple simulations are included.

UR - http://www.scopus.com/inward/record.url?scp=33749852544&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33749852544&partnerID=8YFLogxK

U2 - 10.1007/11840930_36

DO - 10.1007/11840930_36

M3 - Conference contribution

AN - SCOPUS:33749852544

SN - 3540388710

SN - 9783540388715

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

SP - 350

EP - 360

BT - Artificial Neural Networks, ICANN 2006 - 16th International Conference, Proceedings

T2 - 16th International Conference on Artificial Neural Networks, ICANN 2006

Y2 - 10 September 2006 through 14 September 2006

ER -

Discriminant analysis by a neural network with mahalanobis distance

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this