| BP神经网络预测算法性能的改进策略 |
| |
| 引用本文： | 王建平,郭尚. BP神经网络预测算法性能的改进策略[J]. 微电子学与计算机, 2007, 24(10): 144-145,149 |
| |
| 作者姓名： | 王建平 郭尚 |
| |
| 作者单位： | 1. 北京北大方正电子有限公司,北京,100085
2. 北京用友政务软件有限公司,北京,100094 |
| |
| 摘 要： | 预测可以为系统的正常运行提供良好的保障.依据BP神经网络预测算法的特性及电信网络的特点,提出了通过加入周期算法、数据整体上移或下移的预处理来提高预测性能的策略.并对此两种改进策略进行了试验论证.
|
| 关 键 词： | 神经网络 BP神经网络 误差 平均误差 |
| 文章编号： | 1000-7180（2007）10-0144-02 |
| 修稿时间： | 2007-06-02 |
Improvement Strategies of the BP Neural Network Algorithm Performance |
| |
| WANG Jian-ping,GUO Shang. Improvement Strategies of the BP Neural Network Algorithm Performance[J]. Microelectronics & Computer, 2007, 24(10): 144-145,149 |
| |
| Authors: | WANG Jian-ping GUO Shang |
| |
| Affiliation: | 1 Beijing Founder Electronics Co., Ltd, Beijing 100085, China; 2 Beijing UFIDA Co., Ltd, Beijing 100049, China |
| |
| Abstract: | Forecast can give a good protection for the system normal running. In this paper, give an improve forecasting strategies by adding to the cycle of algorithms and data size of the overall upward or downward base on artificial neural network to predict the characteristics and features of telecommunication networks, and do a test to prove the right of the two strategies. |
| |
| Keywords: | neural networks BP neural networks error average error |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
