Some improved methods to analysis stability of recurrent neural networks with interval time-varying delays
| Autor: | Gong-zhi Yu, Kaiyan Zhu, Weibo Song, Kewei Cai, Fengjiao Jiang |
|---|---|
| Rok vydání: | 2016 |
| Předmět: |
0209 industrial biotechnology
Lemma (mathematics) Mathematical optimization Applied Mathematics Activation function Stability (learning theory) Orthogonal complement 02 engineering and technology Interval (mathematics) Computer Science Applications 020901 industrial engineering & automation Recurrent neural network Computational Theory and Mathematics Control theory Convex optimization 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Segmentation Mathematics |
| Zdroj: | International Journal of Computer Mathematics. 94:1228-1251 |
| ISSN: | 1029-0265 0020-7160 |
| Popis: | This paper considers the delay-dependent stability problem of recurrent neural networks with interval time-varying delays. An appropriate Lyapunov–Krasovskii functional is constructed and the combination method of Wirtinger inequality and reciprocally convex optimization technique is employed. Combing a new activation function segmentation method of the boundary condition and the orthogonal complement lemma, three further improved delay-dependent stability criteria are established. Finally, two numerical examples show the effectiveness of our proposed method by comparison with the recent existing works. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|