欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对具有波动性的股票指数预测问题,利用具有强非线性映射能力的广义回归神经网络构建基于历史日（2002年1月7日至2016年6月22日）收盘价、最低价、最高价、成交量、成交额和涨跌幅的股指预测模型,并且在网络的训练过程中利用交叉验证的方式确定了基于均方误差最小的Spread,最后分别从预测误差,预测相对误差等角度对比分析了GRNN神经网络与RBF神经网络的预测精度,得出GRNN神经网络可以很好地实现股指预测的结论.
英文摘要：
In view of the volatility of the stock index prediction,this paper,firstly by applying the generalized regression neural network which has strong nonlinear mapping ability,constructs a stock index prediction model based on the closing price,the lowest price,the highest price,trading volume,turnover and the rise and fall of price during the period from January 7,2002 to June 22,2016.Secondly,in the process of online training,this paper establishes the Spread based on minimum mean square error（MSE）by using cross validation.Thirdly,it compares and analyzes the prediction accuracy of GRNN neural network and RBF neural network from the perspective of the prediction error and the relative prediction error,it is concluded that GRNN neural network can realize the index prediction well.
同期刊论文项目
同项目期刊论文
期刊信息
位置：