Fuzzy seasonal time series for forecasting the production value of the mechanical industry in Taiwan

Based on the seasonal time series ARIMA(p,d,q)(P,D,Q)(s) model (SARIMA) and fuzzy regression model, we combine the advantages of two methods to propos e a procedure of fuzzy seasonal time series and apply this method to foreca sting the production value of the mechanical industry in Taiwan. The intent ion of the article is to provide the enterprises, in this era of diversifie d management, with a fresh method to conduct short-term prediction for the future in the hope that these enterprises can perform more accurate plannin g. This method includes interval models with interval parameters and provid es the possibility distribution of future value. From the results of practi cal application to the mechanical industry, it can be shown that this metho d makes good forecasts. Further, this method makes it possible for decision makers to forecast the possible situations based on fewer observations tha n the SARIMA model and has the basis of pre-procedure for fuzzy time series . (C) 1999 Elsevier Science Inc.