Prediction of household food retail prices based on ARIMA Model

Predictions of retail prices of household foods bear positive meanings in enhancing the living standard of residents. In our paper, first we subdivided the 41 kinds of household food into 5 subcategories basing on trends of their price movements during 2010 to 2011: Smoothly Rising, Rising with Fluctuations, Stable, Horizontal Fluctuating and Concave. Next, we structured separate ARIMA models for each subcategory, and applied such models to predict the prices of the respective subcategories in April and May of 2011. The results suggest that the ARIMA models can produce good simulations and predictions of the retail prices of foods: that the prices of all subcategories of household foods basically preserve an upward trend. The paper concludes with advice on relative policies by the author based on the results obtained, with reference values to the concerned government divisions and relevant parties.