Analysis of ℒ 2-ℒ stability for multilayer hopfield neural networks

Research output: Contribution to journalArticle

Abstract

In this paper, we propose some new conditions on ℒ 2 - ℒ stability of multilayer Hopfield neural networks. These sufficient conditions are represented based on matrix norm and linear matrix inequality (LMI). Under these conditions, multilayer Hopfield neural networks reduce the effect of external input on the state vector to a predefined level. Moreover, the proposed conditions ensure asymptotic stability for multilayer Hopfield neural networks without external input. Dynamic Publishers, Inc.

Original languageEnglish
Pages (from-to)111-118
Number of pages8
JournalNeural, Parallel and Scientific Computations
Volume20
Issue number1
Publication statusPublished - 2012 Mar 1
Externally publishedYes

Fingerprint

Multilayer Neural Network
Hopfield neural networks
Hopfield Neural Network
Multilayer neural networks
Matrix Norm
Asymptotic stability
Linear matrix inequalities
Asymptotic Stability
Matrix Inequality
Linear Inequalities
Sufficient Conditions

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software
  • Computer Networks and Communications
  • Applied Mathematics
  • Theoretical Computer Science

Cite this

Analysis of ℒ 2-ℒ stability for multilayer hopfield neural networks. / Ahn, Choon Ki.

In: Neural, Parallel and Scientific Computations, Vol. 20, No. 1, 01.03.2012, p. 111-118.

Research output: Contribution to journalArticle

@article{2abc9365758d4d1492c83a4021807356,
title = "Analysis of ℒ 2-ℒ ∞ stability for multilayer hopfield neural networks",
abstract = "In this paper, we propose some new conditions on ℒ 2 - ℒ ∞ stability of multilayer Hopfield neural networks. These sufficient conditions are represented based on matrix norm and linear matrix inequality (LMI). Under these conditions, multilayer Hopfield neural networks reduce the effect of external input on the state vector to a predefined level. Moreover, the proposed conditions ensure asymptotic stability for multilayer Hopfield neural networks without external input. Dynamic Publishers, Inc.",
author = "Ahn, {Choon Ki}",
year = "2012",
month = "3",
day = "1",
language = "English",
volume = "20",
pages = "111--118",
journal = "Neural, Parallel and Scientific Computations",
issn = "1061-5369",
publisher = "Dynamic Publishers",
number = "1",

}

TY - JOUR

T1 - Analysis of ℒ 2-ℒ ∞ stability for multilayer hopfield neural networks

AU - Ahn, Choon Ki

PY - 2012/3/1

Y1 - 2012/3/1

N2 - In this paper, we propose some new conditions on ℒ 2 - ℒ ∞ stability of multilayer Hopfield neural networks. These sufficient conditions are represented based on matrix norm and linear matrix inequality (LMI). Under these conditions, multilayer Hopfield neural networks reduce the effect of external input on the state vector to a predefined level. Moreover, the proposed conditions ensure asymptotic stability for multilayer Hopfield neural networks without external input. Dynamic Publishers, Inc.

AB - In this paper, we propose some new conditions on ℒ 2 - ℒ ∞ stability of multilayer Hopfield neural networks. These sufficient conditions are represented based on matrix norm and linear matrix inequality (LMI). Under these conditions, multilayer Hopfield neural networks reduce the effect of external input on the state vector to a predefined level. Moreover, the proposed conditions ensure asymptotic stability for multilayer Hopfield neural networks without external input. Dynamic Publishers, Inc.

UR - http://www.scopus.com/inward/record.url?scp=84859035487&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84859035487&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:84859035487

VL - 20

SP - 111

EP - 118

JO - Neural, Parallel and Scientific Computations

JF - Neural, Parallel and Scientific Computations

SN - 1061-5369

IS - 1

ER -