基于ARIMA、BP神经网络与GM的组合模型

为了提高预测模型的精度,给出了一种新的组合预测模型.利用时间序列ARIMA预测模型、BP神经网络及GM灰色预测模型进行单一模型的拟合与预测,通过赋予适当权系数结合三种方法得到了新的组合预测模型.山西省人均GDP预测实例应用结果表明:组合预测模型很好地描述了山西省人均GDP的非线性发展,比单一预测方法具有更高的预测精度.组合模型发挥了这三种模型各自的优势,可以作为人均GDP预测的有效方法,该模型在时间序列的预测中是有效的.

Combination model based on ARIMA,BP neural network and GM

In order to improve the forecast accuracy of prediction model,this paper presents a new method of the combined forecasting model.By using ARIMA model,the BP neural network and GM grey model to fit and predict time series data respectively,a new combined forecasting method is obtained based on the combination of three methods with reasonable weight coefficients.The forecasting results show that the combined forecasting model can describe the nonlinear development on the per capita GDP in Shanxi Province,of which the prediction accuracy is higher than that of each single prediction model.The forecasting model combines the three models’ respective advantages,which enable the combination model proposed to be an effective model for predicting the per capita GDP.It is also effective in the prediction of time series data.