从理论上讨论了多维动态信用评估(MDCE)的状态空间结构,提出信用状态维度、信用状态空间(CSS)、ε-等信用状态和ε-伪等信用状态等概念.在构造信用状态空间的基础上,结合空间分析方法和优化技术,构建了偏序结构下的多目标评估模型,所有待评估的信用主体被描述为信用状态空间中点的集合,并归入相应的信用状态等级,最后给出一个应用示例.
In this paper, the space frame of multidimensional dynamic credit evaluation (MDCE) problem is discussed theoretically. We put forth the conceptions of credit dimensions, credit statues space (CSS), ε-identical credit status and ε-pseudo identical credit status etc. Combining space conformational analysis, an evaluation multiobjects model on partial ordering is also founded depending on optimization techniques. By using our method, the credit qualities of all evaluated objects can be described as corresponding points in credit space S, and ascribed to a suitable credit rating. At last, a demonstration is given.