| 优化神经元激活函数类型的重要性 |
| |
| 引用本文： | 管建池,肇莹,刘红星.优化神经元激活函数类型的重要性[J].济南大学学报(自然科学版),2005,19(3):282-282. |
| |
| 作者姓名： | 管建池 肇莹 刘红星 |
| |
| 作者单位： | 南京大学,电子科学与工程系,江苏,南京,210093;南京大学,电子科学与工程系,江苏,南京,210093;南京大学,电子科学与工程系,江苏,南京,210093 |
| |
| 基金项目： | 国家自然科学基金资助项目(60275041,59905011) |
| |
| 摘 要： | 以典型的前馈网络设计为例，对网络学习中神经元激活函数类型优化的重要性做了进一步的探讨。理论分析和实际算例显示：神经元激活函数类型相当于信号分解中的基函数，如果不对它进行优化，就不能获得理想的泛化性能。特别是外推泛化性能。
|
| 关 键 词： | 人工神经网络 神经元 激活函数 |
| 文章编号： | 1671-3559(2005)03-0282-01 |
| 修稿时间： | 2005年3月30日 |
Importance of Optimizing Neural Activation Function Types |
| |
| GUAN Jian-chi,ZHAO Ying,LIU Hong-xing.Importance of Optimizing Neural Activation Function Types[J].Journal of Jinan University(Science & Technology),2005,19(3):282-282. |
| |
| Authors: | GUAN Jian-chi ZHAO Ying LIU Hong-xing |
| |
| Abstract: | Perfect artificial neural network (ANN) learning should include the optimization of neural activation function types, and the tradition of optimizing the network weights only in ANN learning is not consistent with biology. As an instance with feed-forward ANN, this paper has further discussed the importance of optimizing neuronal activation function types. With several theoretical analyses and practical examples, it has been shown that like the basis function in signal decomposition, the optimization of it is very important to ANN generalization performance. |
| |
| Keywords: | artificial neural network neuron activation function |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
