DocumentCode
144748
Title
Stability analysis on neural networks with a class of Mexican-hat-type activation functions
Author
Lili Wang
Author_Institution
Dept. of Appl. Math., Shanghai Univ. of Finance & Econ., Shanghai, China
Volume
2
fYear
2014
fDate
26-28 April 2014
Firstpage
1151
Lastpage
1155
Abstract
A class of neural networks with Mexican-hat-type activation functions are discussed in this paper. By differentiating the external inputs Ii (i = 1,2,..., n) into different four ranges, a set of new sufficient conditions are presented to study the multistability of systems. The obtained results improve and extend some recent works.
Keywords
neural nets; stability; transfer functions; Mexican-Hat-type activation functions; neural networks; stability analysis; sufficient conditions; system multistability; Biological neural networks; Cellular neural networks; Indexes; Stability criteria; Trajectory; Mexican-hat activation; Neural networks; multistability;
fLanguage
English
Publisher
ieee
Conference_Titel
Information Science, Electronics and Electrical Engineering (ISEEE), 2014 International Conference on
Conference_Location
Sapporo
Print_ISBN
978-1-4799-3196-5
Type
conf
DOI
10.1109/InfoSEEE.2014.6947851
Filename
6947851
Link To Document