On exponential stability of delayed neural networks with a general class of activation functions

Changyin Sun; Kanjian Zhang; Shumin Fei; Chun Bo Feng

doi:10.1016/S0375-9601(02)00471-1

On exponential stability of delayed neural networks with a general class of activation functions

Changyin Sun
, Kanjian Zhang
, Shumin Fei
, Chun Bo Feng

College of Artificial Intelligence and Robotics

Southeast University, Nanjing

Research output: Contribution to journal › Article › peer-review

74 Scopus citations

Abstract

In this Letter, based on globally Lipschitz continuous activation functions, new conditions ensuring existence, uniqueness and global exponential stability of the equilibrium point of delayed neural networks are obtained. The delayed Hopfield network and bidirectional associative memory network are special cases of the network model considered in this Letter. So this work gives some improvements to the previous ones.

Original language	English
Pages (from-to)	122-132
Number of pages	11
Journal	Physics Letters, Section A: General, Atomic and Solid State Physics
Volume	298
Issue number	2-3
DOIs	https://doi.org/10.1016/S0375-9601(02)00471-1
State	Published - 3 Jun 2002

Keywords

Activation functions
Global exponential stability
M-matrix
Neural networks

Access to Document

10.1016/S0375-9601(02)00471-1

Cite this

@article{ed62968b3bd04f799c121a9ac614315d,

title = "On exponential stability of delayed neural networks with a general class of activation functions",

abstract = "In this Letter, based on globally Lipschitz continuous activation functions, new conditions ensuring existence, uniqueness and global exponential stability of the equilibrium point of delayed neural networks are obtained. The delayed Hopfield network and bidirectional associative memory network are special cases of the network model considered in this Letter. So this work gives some improvements to the previous ones.",

keywords = "Activation functions, Global exponential stability, M-matrix, Neural networks",

author = "Changyin Sun and Kanjian Zhang and Shumin Fei and Feng, \{Chun Bo\}",

year = "2002",

month = jun,

day = "3",

doi = "10.1016/S0375-9601(02)00471-1",

language = "英语",

volume = "298",

pages = "122--132",

journal = "Physics Letters, Section A: General, Atomic and Solid State Physics",

issn = "0375-9601",

publisher = "Elsevier B.V.",

number = "2-3",

}

TY - JOUR

T1 - On exponential stability of delayed neural networks with a general class of activation functions

AU - Sun, Changyin

AU - Zhang, Kanjian

AU - Fei, Shumin

AU - Feng, Chun Bo

PY - 2002/6/3

Y1 - 2002/6/3

N2 - In this Letter, based on globally Lipschitz continuous activation functions, new conditions ensuring existence, uniqueness and global exponential stability of the equilibrium point of delayed neural networks are obtained. The delayed Hopfield network and bidirectional associative memory network are special cases of the network model considered in this Letter. So this work gives some improvements to the previous ones.

AB - In this Letter, based on globally Lipschitz continuous activation functions, new conditions ensuring existence, uniqueness and global exponential stability of the equilibrium point of delayed neural networks are obtained. The delayed Hopfield network and bidirectional associative memory network are special cases of the network model considered in this Letter. So this work gives some improvements to the previous ones.

KW - Activation functions

KW - Global exponential stability

KW - M-matrix

KW - Neural networks

UR - https://www.scopus.com/pages/publications/0037013907

U2 - 10.1016/S0375-9601(02)00471-1

DO - 10.1016/S0375-9601(02)00471-1

M3 - 文章

AN - SCOPUS:0037013907

SN - 0375-9601

VL - 298

SP - 122

EP - 132

JO - Physics Letters, Section A: General, Atomic and Solid State Physics

JF - Physics Letters, Section A: General, Atomic and Solid State Physics

IS - 2-3

ER -

On exponential stability of delayed neural networks with a general class of activation functions

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this