基于案例推理方法建立决策模型

The Establishment of Decision Model Based on CBR(Case-based Reasoning)

解决决策问题需要耗费大量的时间、金钱和精力,最后生成模型,这个模型只适用于一个特定的问题,不具有普适性。为此,提出一种基于案例推理的方法来建立拓扑水平影响图,用以前建模解决决策问题的经验中得到的先验知识,来解决其他类似的决策问题。此外,还提出基于案例的决策类分析(CB-DCA),一种基于案例的推理,打造影响图的方法。CB-DCA由一个案例的检索过程和适应过程组成。有关检索过程有两个方法,一个是拟合比率,另一个是垃圾比。适应过程基于决策分析的知识和决策参与者的特定域的知识。

To solve the decision problem requires a lot of time, money and energy to finally generate a model that is only suitable for a particular problem and doesn＇t have universality. Therefore, a method based on CBR to establish the topological level influence diagram is put forward, which uses the prior knowledge obtained from the previous modeling experiences of solving the decision problem to solve other similar decision problems, besides, a case based-decision class analysis （CB-DCA）,which is a reasoning based on case and a method for creating the influence diagram, is also put forward. CB-DCA consists of retrieval process and adaptation process. There are two methods in retrieval process, one is fitting ratio, and another is refuse porosity ratio. And the adaptation process id based on the knowledge of decision analysis and the specific domain knowledge of decision participators.