| 引用本文: | 李明国,郁文贤.神经网络的函数逼近理论.[J].国防科技大学学报,1998,20(4):70-76.[点击复制] |
| Li Mingguo,Yu Wenxian.On Neural Networks Founction Approximation[J].Journal of National University of Defense Technology,1998,20(4):70-76[点击复制] |
|
| |
|
|
| 本文已被：浏览 6527次 下载 7732次 |
| 神经网络的函数逼近理论 |
| 李明国, 郁文贤 |
|
(国防科技大学 电子技术系 湖南 长沙 410073)
|
| 摘要: |
| 分析了将函数逼近理论与方法引入神经网络研究的必要性; 从经典函数逼近与统计分析两方面详细地讨论了多层前馈网(MLP)逼近能力分析的基本方法及结论; 分析了正则理论观点下的径向基函数网络(RBF)的逼近能力; 讨论了RBF网与多层前馈网在最佳逼近特性上的差异。文末指出了神经网络函数逼近的发展方向。 |
| 关键词: 神经网络, 函数逼近, 正则理论, 推广能力 |
| DOI： |
| 投稿日期：1998-03-02 |
| 基金项目: |
|
| On Neural Networks Founction Approximation |
| Li Mingguo, Yu Wenxian |
|
(Department of Electronics Engineering, NUDT, Changsha, 410073)
|
| Abstract: |
| In this paper, an analysis of the necessity to introduce classical function approximation theory and methods into the research of ANN is presented with a brief introduction of its contents. Then, the main results about the approximation capability of MLP network and the basic analysis methods are detailed from the two aspects of classical function appoximantion and statistical analysis. The approximation capability of RBF ntework is analyzed under the view of Regularity Theory, and the difference of the best approximation property between the RBF and MLP network is revealed. |
| Keywords: artificial neural networks, function approximation, regularization theory, generalization property |
|
|
|
|
|
