An Intellectual Forecasting Support System

Abstract Good forecast which is necessary to all kinds of effective decision can only be obtained by the help of an efficacious forecasting support system. The system studied by us is an intellectual forecasting support system (IFSS) which can be used by both experts in forecasting and common users who want to get a good forecast about something concerned without any knowledge and experience of forecasting technology. The functions of IFSS are: to choose the appropriate forecasting model for users to cope with the event to be forecasted; to build the time series model automatically if this type of model has been selected; to improve the system at any time by adding new models extending model base, deleting, updating old ones and other operations: to examine the forecasting results and choose the best one from them. In this paper, the architecture and framework of IFSS are described and explained, the functional structure of IFSS is introduced, especially, automatic modeling subsystem, intellectual model selection and model management subsystems are emphasized, and then the implementation of IFSS and the use of software engnieering theory are discussed.