| 基于改进BP算法的商品销售量的预测 |
| |
| 引用本文： | 秦绪伟,董傲霜,洪智勇.基于改进BP算法的商品销售量的预测[J].沈阳理工大学学报,2004,23(4):83-86. |
| |
| 作者姓名： | 秦绪伟 董傲霜 洪智勇 |
| |
| 作者单位： | 沈阳理工大学,信息科学与工程学院,辽宁,沈阳,110168 |
| |
| 摘 要： | 提出了一种改进的BP算法，该方法通过结合Cauchy训练来改进传统BP算法，避免传统BP算法容易陷入局部极小点，提高Cauchy训练的训练速度和解决不收敛的问题，并运用该方法于商品销售量的预测，实例表明该方法使网络具有较快的收敛速度和较高的准确度．
|
| 关 键 词： | 神经网络算法 商品销售量 预测方法 模拟退火算法 节点 |
| 文章编号： | 1003-1251(2004)04-0083-04 |
| 修稿时间： | 2004年7月9日 |
Application and Research on Sales Volume Prediction of Commodities Based on Improved BP Algorithm |
| |
| QIN Xu-wei,DONG Ao-shuang,Hong Zhi-yong.Application and Research on Sales Volume Prediction of Commodities Based on Improved BP Algorithm[J].Transactions of Shenyang Ligong University,2004,23(4):83-86. |
| |
| Authors: | QIN Xu-wei DONG Ao-shuang Hong Zhi-yong |
| |
| Abstract: | The disadvantage of the traditional BP algorithm and Cauchy training is analyzed and an improved BP algorithm is put forward. The algorithm improves traditional BP algorithm through integrating Cauchy training to avoid the defect that the traditional BP algorithm falls into local minimum easily. Meanwhile it can avoid the defect that Cauchy training possibly do not converge and its speed is slow. The algorithm is applied to predict sales volume of commodities and the given example proves that the algorithm makes artificial neural network have faster convergence speed and higher accuracy. |
| |
| Keywords: | BP algorithm simulated annealing algorithm Cauchy training sales volume prediction |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
