Abstract
In this paper, two forecasting models based on the grey theory and seasonal index are proposed to deal with seasonal time series forecasting problems. A representative seasonal index (RSI) is used to represent the seasonality in presented models. In addition, the possible variant intervals of the seasonality are obtained based on the RSL The fuzzy seasonality forecasting model, the grey model, and the grey model with simple ratio-to-moving-average model are used to compare the performance of forecasting. Experimental results show that the proposed models outperform the other models in terms of forecasting accuracy.