基于时间序列的多因素耦合下GDP预测研究——以伊犁州为例

Evaluation of GDP Multi Factor Coupling Based on Time-series Analysis——in YiLi prefecture countries

以时间序列分析相关理论为基础，首先介绍了ARIMA模型和VAR模型的建模过程，然后构造ARIMA模型对伊犁州直GDP进行预测，再考虑进出口总额、社会消费品零售额、工业和建筑业、社会固定资产总额等因素的耦合效应，构造VAR模型对伊犁州直GDP进行预测，最后对两种方法做对比分析，结果显示，VAR模型预测精度较高，误差较小，经济含义明确，预测的GDP结果可为伊犁州党委、政府制定相关经济政策和发展战略提供一定的科学依据．

This paper first develops the ARIMA mathematical model to predict the GDP of Yili based on the time series analysis. Then we consider the coupling effect of total imports and exports volume, retail sales of social consumer goods, industrial and construction,the total investment in fixed assets, and construct the VAR model to predict the GDP of Yili. We then compare these two models from the data. The results show that the precision of the VAR model is higher than the other,and the meaning of economy is clear. The prediction of GDP can provide a scientific basis for the relevant economic policies and development strategies of the government of Yili.