Constructively learning a near-minimal neural network architecture

Justin Fletcher; Zoran Obradovic

doi:10.1109/icnn.1994.374163

Constructively learning a near-minimal neural network architecture

Justin Fletcher, Zoran Obradovic

Medical Informatics and Clinical Epidemiology

Research output: Contribution to conference › Paper › peer-review

3 Scopus citations

Abstract

Rather than iteratively manually examining a variety of pre-specified architectures, a constructive learning algorithm dynamically creates a problem-specific neural network architecture. Here we present an revised version of our parallel constructive neural network learning algorithm which constructs such an architecture. The three steps of searching for points on separating hyperplanes, determining separating hyperplanes from separating points and selecting separating hyperplanes generate a near-minimal architecture. As expected, experimental results indicate improved network generalization.

Original language	English (US)
Pages	204-208
Number of pages	5
DOIs	https://doi.org/10.1109/icnn.1994.374163
State	Published - 1994
Event	Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7) - Orlando, FL, USA Duration: Jun 27 1994 → Jun 29 1994

Other

Other	Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7)
City	Orlando, FL, USA
Period	6/27/94 → 6/29/94

ASJC Scopus subject areas

Software

Access to Document

10.1109/icnn.1994.374163

Cite this

@conference{4f63d8dcbc694b799f4b91ee4c024b25,

title = "Constructively learning a near-minimal neural network architecture",

abstract = "Rather than iteratively manually examining a variety of pre-specified architectures, a constructive learning algorithm dynamically creates a problem-specific neural network architecture. Here we present an revised version of our parallel constructive neural network learning algorithm which constructs such an architecture. The three steps of searching for points on separating hyperplanes, determining separating hyperplanes from separating points and selecting separating hyperplanes generate a near-minimal architecture. As expected, experimental results indicate improved network generalization.",

author = "Justin Fletcher and Zoran Obradovic",

note = "Copyright: Copyright 2020 Elsevier B.V., All rights reserved.; Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7) ; Conference date: 27-06-1994 Through 29-06-1994",

year = "1994",

doi = "10.1109/icnn.1994.374163",

language = "English (US)",

pages = "204--208",

}

TY - CONF

T1 - Constructively learning a near-minimal neural network architecture

AU - Fletcher, Justin

AU - Obradovic, Zoran

PY - 1994

Y1 - 1994

N2 - Rather than iteratively manually examining a variety of pre-specified architectures, a constructive learning algorithm dynamically creates a problem-specific neural network architecture. Here we present an revised version of our parallel constructive neural network learning algorithm which constructs such an architecture. The three steps of searching for points on separating hyperplanes, determining separating hyperplanes from separating points and selecting separating hyperplanes generate a near-minimal architecture. As expected, experimental results indicate improved network generalization.

AB - Rather than iteratively manually examining a variety of pre-specified architectures, a constructive learning algorithm dynamically creates a problem-specific neural network architecture. Here we present an revised version of our parallel constructive neural network learning algorithm which constructs such an architecture. The three steps of searching for points on separating hyperplanes, determining separating hyperplanes from separating points and selecting separating hyperplanes generate a near-minimal architecture. As expected, experimental results indicate improved network generalization.

UR - http://www.scopus.com/inward/record.url?scp=0028740110&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028740110&partnerID=8YFLogxK

U2 - 10.1109/icnn.1994.374163

DO - 10.1109/icnn.1994.374163

M3 - Paper

AN - SCOPUS:0028740110

SP - 204

EP - 208

T2 - Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7)

Y2 - 27 June 1994 through 29 June 1994

ER -

Constructively learning a near-minimal neural network architecture

Abstract

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this