| 基于遗传算法的人工神经网络优化设计 |
| |
| 引用本文： | 徐红.基于遗传算法的人工神经网络优化设计[J].燕山大学学报,2004,28(4):337-340. |
| |
| 作者姓名： | 徐红 |
| |
| 作者单位： | 燕山大学,电气工程学院,河北,秦皇岛,066004 |
| |
| 摘 要： | 提出遗传算法新的编码方案，用于全局优化神经网络拓扑结构和权值参数，解决了神经网络拓扑结构的难确定性和权值训练的长时性问题，并且在遗传操作中采用自适应代沟的替代策略，改善其求解效率，获得了良好的优化结果。通过仿真实验显示了该算法的快速性和有效性。
|
| 关 键 词： | 遗传算法 权值 人工神经网络 拓扑结构 显示 快速性 自适应 全局优化 求解 遗传操作 |
| 文章编号： | 1007-791X(2004)04-0337-04 |
| 修稿时间： | 2004年3月28日 |
Artificial neural network's optimal design based on genetic algorithms |
| |
| XU Hong
. College of Electrical Engineering,Yanshan University,Qinhuangdao,Hebei ,China.Artificial neural network''''s optimal design based on genetic algorithms[J].Journal of Yanshan University,2004,28(4):337-340. |
| |
| Authors: | XU Hong
College of Electrical Engineering Yanshan University Qinhuangdao Hebei China |
| |
| Affiliation: | XU Hong
1. College of Electrical Engineering,Yanshan University,Qinhuangdao,Hebei 066004,China |
| |
| Abstract: | A new coding scheme of genetic algorithm is proposed to optimize the topology and weights distribution of neural network, which solves the problems of topology's uncertainty and long-time of weights training. And the strategy of adaptive generation replace is used in genetic operation. It improves efficiency of solution, and a good optimal result is obtained. Simulation shows that this algorithm has speediness and effectiveness. |
| |
| Keywords: | neural network genetic algorithm adaptive generation replace optimize |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
