An application of fuzzy BP neural network for decision support system

A practical decision support system aims at providing decision-makers with intelligent service on the basis of analyzing the current conditions. Such a system must therefore combine techniques for memorizing, learning and processing information with methods of approximate reasoning. In the paper, we put forward a four-layered fuzzy neural network model for an intelligent decision support system (IDSS). The contribution of the paper is the design of a hierarchy neural network and three kinds of fuzzy information process neurons. The novel model is able to realize fuzzy mapping, fuzzy reasoning and decision making according to the initial information relative to the problem. The emphasis is placed on describing the construction, characteristics and reasoning mechanism of the four-layered fuzzy neural network. Furthermore, we analyze the feasibility and rationality of the model as an IDSS in theory.