Decision-making Based On Interval Concept Lattice Model Construction And Optimization Principle

Under the rapid development of information science, policymakers are faced with the situation that the information is too much, dynamic and even noisy. The demand for the technology of intelligent decision is becoming more and more urgent, and the decision technology has not been confined to the traditional statistics and forecast any more. However,the application of knowledge discovery and association found is popular recently. The association rules extracted from interval concept lattice is applied to the process of intelligent decision reasoning. Firstly, the characteristics and properties of interval concept lattice were analyzed, and the association rules of interval concept lattice according to the relevance of lattice structure were extracted. Secondly, the association rules extracted from interval concept lattice for reasoning and decision-making were used. Association rules based on interval concept lattice are similar to deduction, induction and analogy. At the same time, they have their own characteristics: the decision rule is not only one form of revealing data, but also relate to the corresponding data. And, it can be used under uncertain situation. Then,the threshold of interval concept lattice with the method of game theory was optimized. The determination of two parameters of threshold is a method of a game between the accuracy and the generalized. The consequence of a game is satisfied with two sides’ initial required value.At last, an example based on the partial data of data library of dermatology from UCI data sets was made, and then the decision reasoning process in medical diagnosis was analyzed.