|
|||||
|
|
| 三种生成神经网络拓扑结构的方法在股票、商品价格预测中的应用及结果比较 | |
| 引用本文： | 刘畅,张文,邵燕敏.三种生成神经网络拓扑结构的方法在股票、商品价格预测中的应用及结果比较[J].系统科学与数学,2011,31(3). |
| 作者姓名： | 刘畅 张文 邵燕敏 |
| 作者单位： | 1. 中国科学院数学与系统科学研究院,北京,100190 2. 中国科学院虚拟经济与数据科学研究中心,北京,100039 3. 中国科学院数学与系统科学研究院,北京100190;中国科学院预测科学研究中心,北京100190 |
| 基金项目： | 中国科学院重大方向性项目“经济监测预警及政策模拟仿真平台预研项目”(编号KACX1-YW-0906). |
| 摘 要： | 对比了三种不同神经网络模型的生成方式:传统神经网络生成模型,遗传算法训练神经网络模型,以及在第二种方式训练参数的基础上,再使用传统神经网络优化生成模型.论文使用上述三种方法对代表性股票和商品价格进行拟合并预测,通过预测结果准确性和稳定性的比较发现:引入遗传算法后的神经网络在样本内的拟合误差有所降低,而第三种方法在样本外有最低的预测误差和最优稳定性. |
| 关 键 词： | 遗传算法 神经网络 股票 商品 价格预测 |
THE DIFFERENT METHODS OF THE COMBINATION OF GENETIC ALGORITHM AND NEURAL NETWORK AND THEIR APPLICATION ON THE PRICE PREDICTION OF STOCK AND COMMODITY |
|
| LIU Chang,ZHANG Wen,SHAO Yanmin.THE DIFFERENT METHODS OF THE COMBINATION OF GENETIC ALGORITHM AND NEURAL NETWORK AND THEIR APPLICATION ON THE PRICE PREDICTION OF STOCK AND COMMODITY[J].Journal of Systems Science and Mathematical Sciences,2011,31(3). | |
| Authors: | LIU Chang ZHANG Wen SHAO Yanmin |
| Institution: | LIU Chang (Academy of Mathematics and Systems Science,Chinese Academy of Sciences,Beijing 100190) ZHANG Wen (CAS Research Center on Fictitious Economy(?) Data Science,Management School of Chinese Academy of Sciences 100139) SHAO Yanmin (Academy of Mathematics and Systems Science,Beijing 100190,Center for Forecasting Science,Chinese Academy of Sciences Beijing 100190) |
| Abstract: | Three methods of the parameter formation of the neural network model are compared:The traditional neural network training,the genetic algorithm substititing for the training process,and based on the second method,training the network using the parameters optimized from the genetic algorithm responsively.All the three methods are applied on the prediction of the stock and commodity prices,and from the analyzing of the veracity and the stability of the prediction result,it is easy to see that after the introd... |
| Keywords: | wavelet neural network genetic algorithm stock commodity price prediction |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |