不完备序区间值决策系统中的可信规则获取

以不完备序区间值决策系统为研究对象,其中不仅包含遗漏型未知区间值,而且属性值域为全序集.给出了未知区间值的三种形式及其填充式区间值的定义,引入灰的白化方法用以构建一个新的填充式不完备序白化值决策系统,并讨论其在优势和弱势关系下的可信规则获取.进一步研究了优势和弱势对象的约简以及其决策类的相对约简问题,给出了相应的判定定理与区分函数,为最终从不完备序区间值决策系统中获取最优可信决策规则提供了新的理论基础与操作手段.、

Acquisition of Credible Rules in Incomplete Ordered Interval-valued Decision Systems

In this paper,an incomplete ordered decision system for interval value processing is deeply investigated,in which not only unknown interval value are existing but also all attributes are looked as criterions.In such information system,the concepts of three types of unknown interval value and fill-in interval value are given;Furthermore,we introduce a method of grey's whitening in fill-in incomplete ordered interval-valued decision system and construct a new ordered whitening value decision system,from which we can obtain the credible rules of the increasing and decreasing objects.We investigate the reductions and relative reductions in terms of the decision classes of these two objects,the judgment theorems and discernibility functions with respect to these reductions are obtained,from which we have obtained new theory and practical approach to generate all optimal credible decision rules from the incomplete ordered interval-valued decision system.