A two-stage rule base optimization method based on combination of classical algorithm and GA for intelligent decision support system

Intelligent decision support system performance depends on knowledge base quality. With the decision problem and knowledge base becoming more complicated, it is necessary to provide effective method to optimize the management of knowledge base. Based on detailed analyzing of running characteristics and potential detects of rule base in IDSS, a novel two-stage optimization method is proposed. Conventional optimization approach and genetic algorithm are combined to recognize and eliminate kinds of defects in rule base. Exact defect definitions and optimization operation details are given. The sample shows that the method is feasible in practice.