| 梯度网络平稳点集合的Hopfield稳定性与吸引域 |
| |
| 引用本文： | 冯芙叶,刘志勤,聂成. 梯度网络平稳点集合的Hopfield稳定性与吸引域[J]. 长安大学学报(自然科学版), 2003, 23(3): 107-110 |
| |
| 作者姓名： | 冯芙叶 刘志勤 聂成 |
| |
| 作者单位： | 1. 空军工程大学,导弹学院,陕西,咸阳,738000;西安科技学院,基础课部,陕西,西安,710054
2. 西安科技学院,基础课部,陕西,西安,710054 |
| |
| 摘 要： | 给出梯度神经网络模型的电路实现 ,证明了网络平稳点集合的全局 Hopfield稳定性。对各个平稳点吸引域的几何估计得到了相关结果 ,这为网络初始输入的数量与方向判定提供了基本的量度参数准则。
|
| 关 键 词： | 神经网络 H-稳定性 吸引域 |
| 文章编号： | 1671-8879(2003)03-0107-04 |
| 修稿时间： | 2002-03-10 |
Hopfield-stability of equilibrium point set for gradient neural networks and their attractive regions |
| |
| FENG Fu ye ,,LIU Zhi qin ,NIE Cheng. Hopfield-stability of equilibrium point set for gradient neural networks and their attractive regions[J]. JOurnal of Chang’an University:Natural Science Edition, 2003, 23(3): 107-110 |
| |
| Authors: | FENG Fu ye LIU Zhi qin NIE Cheng |
| |
| Affiliation: | FENG Fu ye 1,2,LIU Zhi qin 2,NIE Cheng 2 |
| |
| Abstract: | Neural networks for problems of the global optimization were proposed and the global convergence of the equilibrium point set was given.The geometric symptoms of the attractive region for each stationary point were analyzed to give a basic rule for choosing the initial values of the proposed neural networks both in qualities and directions. |
| |
| Keywords: | neural networks H stability attractive region |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
