ABSTRACT
The fuzzy exponential smoothing model has been used for short-term forecasting with a minimum of collected data or unknown system structure to determine a better extrapolative interval in a fuzzy set for an unknown future trend. In order to enlarge its applications, this model has been investigated and proposed for solving trend datum and seasonal datum. However, two characters in the fuzzy exponential smoothing model have never been clearly discussed. First, the old smoothed values estimated from the grey model GM(1,1) need further discussion and clarification. Second, the meanings of the tolerance values of the fuzzy constraints in the fuzzy exponential smoothing model require further elucidation. Therefore, in this study, in-depth examinations of fuzzy exponential smoothing model are carried out in order to elucidate the above two problems and to enhance the abilities of the fuzzy exponential smoothing model. An example is given for illustration.