A genetic algorithm-based grey method for forecasting food demand after snow disasters: an empirical study

Determining the food supply after natural disasters is necessary to ensure the safety and social stability of people in disaster areas. An accurate prediction of food demand can help in the creation of a rational food supply program after natural disasters. This study proposes a grey prediction method to deal with irregular fluctuations in food demand after snowstorms. A GM(1,1) model with adaptive background values was established, and the Fourier series was applied to describe the irregular fluctuations in residuals. A genetic algorithm was designed based on GM(1,1) and Fourier series to optimize model parameters and to minimize the mean absolute percentage error. An optimal predictive function was also constructed by using the combined GM(1,1), Fourier series, and optimal parameters. The proposed forecasting method was used to predict three vegetables demand after the 2008 Chinese winter storm and was compared with the traditional GM(1,1) model. Results show that the proposed method has superior forecasting performance over traditional grey methods. Copyright Springer Science+Business Media Dordrecht 2013