| 用BP神经网络实现函数逼近的方法及其讨论 |
| |
| 引用本文： | 张静亚,潘启勇. 用BP神经网络实现函数逼近的方法及其讨论[J]. 常熟理工学院学报, 2004, 18(4): 34-37 |
| |
| 作者姓名： | 张静亚 潘启勇 |
| |
| 作者单位： | 苏州大学电子信息学院,江苏,苏州,215006;常熟理工学院物理系,江苏,常熟,215500;常熟理工学院物理系,江苏,常熟,215500 |
| |
| 摘 要： | 构造一个三层BP神经网络,实现了连续函数的逼近.讨论了函数值在非[0,1]区间时,样本的归一化问题,提出了新的变换关系式,加快了算法的收敛速度.
|
| 关 键 词： | 神经网络 BP算法 归一化 收敛速度 局部最小点 |
| 文章编号： | 1008-2794(2004)04-0034-04 |
| 修稿时间： | 2004-03-20 |
The Realization and Discussion of a Function's Approximation Based on BP Neural Networks |
| |
| ZHANG Jing-ya. The Realization and Discussion of a Function's Approximation Based on BP Neural Networks[J]. Journal of Changshu Institute of Technology, 2004, 18(4): 34-37 |
| |
| Authors: | ZHANG Jing-ya |
| |
| Affiliation: | ZHANG Jing-ya~ |
| |
| Abstract: | This paper introduced a three layer BP neural network,and realized the approximation of a continuous function.Based on the above BP neural network,A new transform normalized the function's output are put forward to and the training speed is fastened,which is proved by the result. |
| |
| Keywords: | Neural Network BP Algorism Normalization Training Speed Function Approximation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
