基于SARIMA模型的安徽省CPI预测

Prediction of CPI in Anhui Province Based on SARIMA Model

收集了安徽省2000年1月至2016年6月的居民消费价格指数(CPI)数据,并借助于统计分析工具Eviews软件进行相关实验。通过CPI的时序图和ADF检验得到该CPI序列为非平稳序列,并且通过序列的相关图,建立了SARIMA时间序列模型,然后通过AIC、SC准则和残差分析确定了最优的预测模型,利用预测模型对安徽省CPI进行了短期预测。实验结果表明:安徽省CPI具有季节性的特征,利用SARIMA模型建模得到的预测模型,总体预测效果较好,且具有一定的现实意义。

The Consumer Price Index （CPI） data of Anhui province from January 2000 to June 2016 was collected in this paper,and the experiment was carried out by means of statistical analysis tool Eviews. Firstly,the CPI sequence was non-stationary sequence that obtained by CPI＇s timing chart and ADF test, and through the correlation graph of the sequence, the SARIMA time series model was established. Then, the optimal prediction model was determined by AIC, SC criterion and residual analysis. Finally, the fore- casting model was used to forecast the CPI of Anhui Province. The experimental results showed that the consumption index of Anhui province has seasonal characteristics, the prediction model has good predic- tion effect by using the SARIMA model, and it has some practical significance.