| 一种新的小波神经网络结构优化设计方法 |
| |
| 引用本文： | 李益国,沈炯,吕震中.一种新的小波神经网络结构优化设计方法[J].控制理论与应用,2003,20(3):329-331. |
| |
| 作者姓名： | 李益国 沈炯 吕震中 |
| |
| 作者单位： | 东南大学动力工程系,江苏,南京,210096 |
| |
| 摘 要： | 针对基于框架的小波神经网络存在的网络冗余性较大的问题, 提出一种基于粗糙集理论的网络结构优化设计方法. 首先通过时频分析确定小波神经网络的初步构造; 在此基础上, 根据网络输出对隐层节点依赖度的大小去除冗余的隐层节点, 达到优化网络结构的目的, 仿真结果表明该方法是简单而有效的.
|
| 关 键 词： | 粗糙集理论 小波神经网络 小波框架 属性依赖度 |
| 文章编号： | 1000-8152(2003)03-0329-03 |
| 收稿时间： | 2001/9/19 0:00:00 |
| 修稿时间： | 2002/5/13 0:00:00 |
New approach to structure optimization of wavelet neural network |
| |
| LI Yi-guo,SHEN Jiong,LU Zhen-zhong.New approach to structure optimization of wavelet neural network[J].Control Theory & Applications,2003,20(3):329-331. |
| |
| Authors: | LI Yi-guo SHEN Jiong LU Zhen-zhong |
| |
| Affiliation: | Department of Power Engineering, Southeast University, Jiangsu Nanjing 210096, China |
| |
| Abstract: | Based on the rough sets theory, an approach was presented to minimize the redundancy of structure existing in wavelet neural networks. The original structure of wavelet network is obtained through time-frequency analysis. Then the redundant nodes are eliminated in light of dependency between the output of the network and the nodes in the hidden layers to optimize the structure of wavelet network. Simulation results illustrate the proposed method is simple and effective. |
| |
| Keywords: | rough sets wavelet neural network wavelet frame dependency of attributes |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《控制理论与应用》浏览原始摘要信息 |
|
点击此处可从《控制理论与应用》下载全文 |
|
