Abstract
The paper considers a general system of ordinary differential equations appearing in the neural network theory. The activation functions are assumed to be continuous and bounded by power type functions of the states and distributed delay terms. These activation functions are not necessarily Lipschitz continuous as it is commonly assumed in the literature. We obtain sufficient conditions for exponential decay of solutions.
| Original language | English |
|---|---|
| Pages (from-to) | 184-190 |
| Number of pages | 7 |
| Journal | Journal of Contemporary Mathematical Analysis |
| Volume | 52 |
| Issue number | 4 |
| DOIs | |
| State | Published - 1 Jul 2017 |
Bibliographical note
Publisher Copyright:© 2017, Allerton Press, Inc.
Keywords
- Neural network
- activation function
- distributed delay
- exponential stabilization
ASJC Scopus subject areas
- Analysis
- Control and Optimization
- Applied Mathematics