On the Effectiveness of GM(1,1) Model with Different Sequence Sizes: An Evidence from Random Simulations

摘要

Various forecasting models have been developed to reveal the real world in a better way. The GM(1,1) model is used to forecast the observations or statistical data which deviate from the true values of the actual system because of disturbances and errors. The model is quite useful in the small sample forecasting. However, each forecasting model has its own application condition. Although there had been some researches on the accuracy and applicable conditions of grey models, these researches mainly focused on model performance for specific case. In this paper, a random simulation method for the accuracy test of GM(1,1) was proposed The results of the random simulation and some real cases demonstrate that the small sample usually has more accuracy than the large sample when establishing grey model in theory.