基于BP神经网络改进的GM(1,N)模型在经济预测中的应用——以河南省经济为例

THE FORECAST AND APPLICATION OF IMPROVED GREY GM(1,N) MODEL BASED ON BP NEURAL NETWORK IN ECONOMY——Taking Henan Province economy as the example

论文就修正GM(1,N)预测模型的误差,提出了新方法。使用BP神经网络对预测模型的残差进行预测,得到的残差预测值对所建模型的预测值进行残差修正,以减少因子变量预测误差对行为变量预测的影响。实践表明这些改进模型可以有效地提高GM(1,N)模型的预测精度。

A new method is proposed to revise the error of GM（ 1,N） forecast model in this paper. The BP neural network is used to carry out the forecast for the forecast model residual,and the obtained the residual predicted value is then applied to the predicted modeling value to revise the residual in order to reduce influence of the factor variable prediction error to the behavior variable forecast. The practice indicates that the improved model may help GM（1,N） model to enhance the forecast precision effectively.