| 改进BP神经网络在经济预测中的应用 |
| |
| 引用本文： | 郇红艳,郇洪江.改进BP神经网络在经济预测中的应用[J].统计与信息论坛,2008,23(1):58-62. |
| |
| 作者姓名： | 郇红艳 郇洪江 |
| |
| 作者单位： | 1. 阜阳师范学院,经济与商业学院,安徽,阜阳,236041
2. 南京工业大学,信息科学与工程学院,江苏,南京,210009 |
| |
| 基金项目： | 国家统计局全国统计科学研究计划重点项目 |
| |
| 摘 要： | 针对传统BP学习算法收敛速度慢、对步长依赖明显等缺点，提出一种利用搜索较优步长的BP算法。其在网络训练中，能够在每次迭代中搜索出一个相对合理的步长，从而使步长的选择对学习速度的影响大大降低。对经济预测仿真结果表明，新算法对步长选择的依赖性小于传统BP算法。
|
| 关 键 词： | 神经网络 BP算法 搜索较优步长 经济预测 |
| 文章编号： | 1007-3116(2008)01-0058-05 |
| 修稿时间： | 2007年7月12日 |
Application of Improved BP Neural Network in Economic Forecasts |
| |
| HUAN Hong-yan,HUAN Hong-jiang.Application of Improved BP Neural Network in Economic Forecasts[J].Statistics & Information Tribune,2008,23(1):58-62. |
| |
| Authors: | HUAN Hong-yan HUAN Hong-jiang |
| |
| Institution: | HUAN Hong-yan, HUAN Hong-jiang (1. School of Economic and Business, Fuyang Normal College, Fuyang 236041, China; 2. School of Information Science and Technology, Nanjing University of Technology, Nanjing 210009, China) |
| |
| Abstract: | The traditional BP algorithm has some limitations such as the slow learning convergence speed and the dependency of the learning process on the learning rate. In view of the limitations, this paper presents a BP neural network based on searching for more adaptive learning rate algorithm. In the network training phase, after each loop the algorithm can find a more adaptive learning rate, so the new algorithm can weaken the negative effect of the choice of learning rate. The simulation results show that the new algorithm is less dependent on the learning rate than the traditional BP algorithm. |
| |
| Keywords: | neural network BP algorithm searching for more adaptive learning rate economic forecasts |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
