Capacity of the upstart algorithm

Ansgar H L West; David Saad

doi:10.1007/978-1-4615-6099-9_65

Capacity of the upstart algorithm

Ansgar H L West, David Saad

Research output: Chapter in Book/Published conference output › Chapter

Abstract

The storage capacity of multilayer networks with overlapping receptive fields is investigated for a constructive algorithm within a one-step replica symmetry breaking (RSB) treatment. We find that the storage capacity increases logarithmically with the number of hidden units <span class='mathrm'>K</span> without saturating the Mitchison-Durbin bound. The slope of the logarithmic increase decays exponentionally with the stability with which the patterns have been stored.

Original language	English
Title of host publication	Mathematics of Neural Networks: Models, Algorithms and Applications
Editors	Stephen W. Ellacott, John C. Mason, Iain J. Anderson
Place of Publication	Oxford
Publisher	Kluwer
Pages	372-378
Number of pages	7
ISBN (Print)	0-7923-9933-1
DOIs	https://doi.org/10.1007/978-1-4615-6099-9_65
Publication status	Published - 1997

Bibliographical note

The original publication is available at www.springerlink.com

Keywords

algorithms
design
experimentation
human factors
measurement
performance
reliability
security
theory
werification

Access to Document

10.1007/978-1-4615-6099-9_65

Capacity of the upstart algorithm
The original publication is available at www.springerlink.com
Accepted author manuscript, 165 KB

Cite this

@inbook{2bee3aedfbd94d60b54419c9af032768,

title = "Capacity of the upstart algorithm",

abstract = "The storage capacity of multilayer networks with overlapping receptive fields is investigated for a constructive algorithm within a one-step replica symmetry breaking (RSB) treatment. We find that the storage capacity increases logarithmically with the number of hidden units K without saturating the Mitchison-Durbin bound. The slope of the logarithmic increase decays exponentionally with the stability with which the patterns have been stored.",

keywords = "algorithms, design, experimentation, human factors, measurement, performance, reliability, security, theory, werification",

author = "West, {Ansgar H L} and David Saad",

note = "The original publication is available at www.springerlink.com",

year = "1997",

doi = "10.1007/978-1-4615-6099-9_65",

language = "English",

isbn = "0-7923-9933-1",

pages = "372--378",

editor = "Ellacott, {Stephen W.} and Mason, {John C.} and Anderson, {Iain J.}",

booktitle = "Mathematics of Neural Networks: Models, Algorithms and Applications",

publisher = "Kluwer",

address = "Netherlands",

}

TY - CHAP

T1 - Capacity of the upstart algorithm

AU - West, Ansgar H L

AU - Saad, David

N1 - The original publication is available at www.springerlink.com

PY - 1997

Y1 - 1997

N2 - The storage capacity of multilayer networks with overlapping receptive fields is investigated for a constructive algorithm within a one-step replica symmetry breaking (RSB) treatment. We find that the storage capacity increases logarithmically with the number of hidden units K without saturating the Mitchison-Durbin bound. The slope of the logarithmic increase decays exponentionally with the stability with which the patterns have been stored.

AB - The storage capacity of multilayer networks with overlapping receptive fields is investigated for a constructive algorithm within a one-step replica symmetry breaking (RSB) treatment. We find that the storage capacity increases logarithmically with the number of hidden units K without saturating the Mitchison-Durbin bound. The slope of the logarithmic increase decays exponentionally with the stability with which the patterns have been stored.

KW - algorithms

KW - design

KW - experimentation

KW - human factors

KW - measurement

KW - performance

KW - reliability

KW - security

KW - theory

KW - werification

UR - https://link.springer.com/chapter/10.1007/978-1-4615-6099-9_65

U2 - 10.1007/978-1-4615-6099-9_65

DO - 10.1007/978-1-4615-6099-9_65

M3 - Chapter

SN - 0-7923-9933-1

SP - 372

EP - 378

BT - Mathematics of Neural Networks: Models, Algorithms and Applications

A2 - Ellacott, Stephen W.

A2 - Mason, John C.

A2 - Anderson, Iain J.

PB - Kluwer

CY - Oxford

ER -

Capacity of the upstart algorithm

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this