中国消费者价格指数预测模型的选择  

Selection of Forecasting Model of China’s Consumer Price Index

作　　者：伊力扎提·艾热提 Elzat Ayrat(School of Economics and Management,Beijing Jiaotong University,Beijing 100044,China)

机构地区：[1]北京交通大学经济管理学院,北京100044

出　　处：《统计与决策》2022年第4期68-73,共6页Statistics & Decision

摘　　要：文章基于R语言分析居民消费价格指数序列,并用指数平滑模型、SARIMA乘法模型、条件异方差模型和协整模型对其进行预测。结果表明,我国居民消费价格指数存在一定的季节性,进行时间序列分析时需考虑季节特性对预测值的影响;通过条件异方差模型消除残差异方差性可以提高预测精度;将商品零售价格指数和流通中的现金同比增长指数作为外在驱动因素,构建协整模型可以进一步保证预测的准确性。This paper analyzes the Consumer Price Index(CPI) series based on R language and predicts it with exponential smoothing model, SARIMA multiplication model, conditional heteroscedasticity model and co-integration model. The results go as below: China’s CPI is of seasonality, and the influence of seasonal characteristics on the predicted value should be considered in time series analysis. The prediction accuracy can be improved by eliminating the heteroscedasticity of residuals with conditional heteroscedasticity model. Using retail price index and year-on-year growth index of cash in circulation as external driving factors to construct co-integration model can further guarantee the accuracy of prediction.

关 键 词：R语言 指数平滑模型 SARIMA乘法模型 条件异方差模型 协整模型 

分 类 号：C813[社会学—统计学] F201