The Intelligent Forecasting System Research

There are forecasting problems everywhere, and the precondition of the right decision is accurate forecasting. Nowadays, enterprises are emphasizing the subtle management day by day, so they are more and more paying attention to forecasting, and forecasting becomes enterprises' daily work. However, there are so many forecasting models which come from statistics,economics,mathematics,sociology as well for forecasting at the present time, which needs forecasting workers understand deeply forecasting models that come from different fields, so that we can make more accurate forecasting. So how to select an adequate model from so many models for forecasting workers becomes a problem that needs to be solved quickly.Traditional forecasting software appeared as a foundational platform of statistical analyses; this software can't meet enterprises' business needs. And forecasting workers must spend too much time and energy to solve problems by using traditional forecasting software. How to design a forecasting system to help enterprises improve the efficiency of their daily forecasting work is another problem which needs to be solved quickly.The paper represented an intelligent forecasting system which based on practical problems, and which is business oriented. What is the intelligent forecasting system? It is used to solve forecasting problem, and makes use of information technology, then helps decision maker to select or combine the best forecasting model from a number of forecasting models and makes it "learn by itself", so that the same problem's forecasting can become more and more accurate. This paper also represents a new method of model base management by using XML and self description technology.