基于ARIMA-SVM模型的郑州市CPI预测研究

Study of Prediction of CPI in Zhengzhou City Based on ARIMA-SVM Model

利用ARIMA-SVM组合预测模型对郑州市的月度居民消费者价格指数(CPI)进行预测。首先,分别用ARIMA模型和SVM预测模型对郑州市月度CPI进行预测;然后,采用ARIMA-SVM组合预测模型对其CPI进行预测,并将结果与单独的ARIMA模型和SVM模型进行比较。实证结果表明,ARIMA-SVM模型能更准确地预测郑州市的CPI指数。

The ARIMA-SVM combined forecasting model is applied to forecast the monthly consumer price index of Zhengzhou. Firstly, the monthly CPI of Zhengzhou City is predicted by ARIMA model and SVM model respectively. Then, the ARIMA-SVM combined forecasting model is used to predict CPI, and the data of which are weighed against the individual ARIMA model and SVM model. The results illustrate that the ARIMA-SVM model combined with the two prediction models can predict the CPI of Zhengzhou more accurately.