Forecasting the CPI Using a Hybrid Sarima and Neural Network Model with Web News Articles

Abstract

Web news fills our life from national affairs to small matters, containing the latent useful information that can reflect the trend of consumer price index. Most previous studies forecast the CPI basing on the historical data while in this paper, the external information is considered and modeled by using the combination of neutral network and seasonal ARIMA model in order to correct the forecasting error for more accurate prediction. The experiments show that the proposed method is better than both the single neutral network and the seasonal ARIMA. The findings imply the web news can bring more precise results in CPI forecasting.