| 高阶反馈型神经网络及其在优化计算中的应用 |
| |
| 引用本文： | 李海波. 高阶反馈型神经网络及其在优化计算中的应用[J]. 东南大学学报(自然科学版), 1990, 20(6): 1-8 |
| |
| 作者姓名： | 李海波 |
| |
| 作者单位： | 东南大学无线电工程系 |
| |
| 摘 要： | 本文用神经网络来求解多元函数的极值.针对 Hopfield 网络的局限,提出了一种高阶反馈型神经网络,并从数学上证明了这个网络具有良好的稳定性.当阶数为2时,功能类似于 Hopfield 网络,但比它稳定.本文还给出求解多元高次函数全局最优解的方法.最后给出了该网络的计算机数值仿真.
|
| 关 键 词： | 神经网络 最优化算法 多项式 |
High-Order Back-Forward Neural Network and Its Application for Optimization Computing |
| |
| Li Haibo. High-Order Back-Forward Neural Network and Its Application for Optimization Computing[J]. Journal of Southeast University(Natural Science Edition), 1990, 20(6): 1-8 |
| |
| Authors: | Li Haibo |
| |
| Affiliation: | Department of Radio Engineering |
| |
| Abstract: | The paper uses Artificial Neural Network to solve optimization problems of polynomial function.In order to overcome the limitation of Hopfield network,a high-order back-forward neural netword is suggested. The network has been proved to be good in numerical stability.We also give two algorithms for solving the global optimization problems of polynomial function. |
| |
| Keywords: | neural network optimization algorithms polynomial |
| 本文献已被 CNKI 维普 等数据库收录！ |
