A forecasting support system based on neural network

This paper presents a new type of forecasting support system because it is very difficult to build an accurate model for some systems, such as the hotel. The forecasting analysis method is based on the radial basis function neural network. The entire forecasting support system data are stored in the database. According to the varies forecasting support subsystem, the synthesized datum and information, which can be taken as the forecasting input samples, are filtered by online analysis processing and data mining. When this method is used in the forecasting support subsystem of the material department of a hotel, it can provide forecasting data for the department manager. With this method the manager can make scientific stocking arrangement. Furthermore, it assures the correct level of materials being stocked in the department for effective running of the hotel