基于ARIMA-DGM-BP组合模型的喀什地区GDP预测

GDP Forecast of Kashi Region Based on ARIMA-DGM-BP Combined Model

组合模型的预测效果一般优于单一模型的预测效果.利用ARIMA模型,DGM(1,1)模型以及BP神经网络构建ARIMR-DGM-BP组合预测模型,给出了建立该组合模型的基本思路.利用喀什地区2000-2018年GDP的相关数据资料,建立了ARIMR-GM-BP和ARIMR-DGM-BP组合模型并对预测的效果进行了统计分析,结果表明ARIMR-DGM-BP组合模型的预测效果优于ARIMR-GM-BP组合模型.最后运用本文的组合预测模型对喀什地区2019-2021年GDP进行预测.

Combined model forecast effect is generally better than that of single model forecast effect.ARIMA model,DGM(1,1)model and artificial neural network BP are used to build the ARIMA-DGM-BP combined forecast model,and the basic idea of establishing the combined forecast model is given.Then,based on relevant data of GDP in Kashi from 2000 to 2018,the combination model of ARIMA-GM-BP and ARIMA-DGM-BP is established and the statistical analysis of the prediction results shows that the prediction effect of ARIMA-DGM-BP is better than that of ARIMA-GM-BP.Finally,this paper takes advantage of the combined model to forecast the GDP of Kashi from 2019 to 2021.