组合时间序列模型及其在我国农业总产值预测中的应用

Prediction of faming gross output in China based on combination time series model

基于组合预测理论,首先建立了我国农业总产值的ARIMA和Holt双参数线性指数平滑单项时间序列预测模型;对模型进行检验后,根据标准差法对各模型进行权重分配,建立我国农业总产值组合预测模型。通过对比证明,组合时间序列模型能在一定程度上克服单项模型缺陷,提高预测精度。

Base on the theory of combination forecasting,ARIMA and Holt were used to establish individual time series forecasting models for farming gross output in China.After testing the models,weights were assigned to each prediction model according to standard deviation method,then a combination forecasting model was established.Experimental results showed that a certain extent of the combination model not only get rid of the defects of the former models,but also raised the accuracy of the prediction.