| 有理式多层前馈神经网络 |
| |
| 引用本文： | 梁久祯,何新贵.有理式多层前馈神经网络[J].控制与决策,2004,19(3):349-350. |
| |
| 作者姓名： | 梁久祯 何新贵 |
| |
| 作者单位： | 1. 浙江师范大学,计算机科学研究所,浙江,金华,321004
2. 北京大学,信息科学与技术学院,北京,100084 |
| |
| 摘 要： | 提出了有理式多层前馈神经网络的数学模型，给出了有理式多层神经网络的学习算法，就计算复杂度而言，有理式神经网络的学习算法与传统的多层神经网络反传播算法是同阶的，函数逼近应用实例结果表明，将有理式多层神经网络用于解决传统问题是有效的。
|
| 关 键 词： | 有理式 神经网络 学习算法 函数逼近 |
| 文章编号： | 1001-0920(2004)03-0349-02 |
Rational fraction multiplayer feed forward neural networks |
| |
| LIANG Jiu-zhen,HE Xin-gui.Rational fraction multiplayer feed forward neural networks[J].Control and Decision,2004,19(3):349-350. |
| |
| Authors: | LIANG Jiu-zhen HE Xin-gui |
| |
| Affiliation: | LIANG Jiu-zhen~1,HE Xin-gui~2 |
| |
| Abstract: | A mathematic model of rational fraction multiplayer feed forward neural networks is proposed. A learning algorithm for rational fraction multiplayer neural networks is presented. The learning algorithm has the same degree of computing complexity as traditional multiplayer neural networks. The function approximation is also discussed. Experiment result illustrates the effectiveness of the rational fraction multiplayer feed forward neural networks in solving traditional problems. |
| |
| Keywords: | rational fraction neural networks learning algorithm function approximation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《控制与决策》浏览原始摘要信息 |
|
点击此处可从《控制与决策》下载全文 |
