不完备信息系统下的属性约简算法

An Algorism of Attributes Reduction in Incomplete Information System

传统的粗糙集模型是处理完全信息系统的有力工具,但对于不完全信息系统却显得无能为力。因此对不完备信息系统的研究也是粗集理论研究领域之一。本文在M.Kryszkiewcz提出的一个容差关系的基础上提出改进,使之更加具有灵活性。然后以该改进模型为基础,运用属性重要性理论,给出属性约简算法,并通过一个实例来验证。最后将该算法和经典算法进行了比较。

Traditional rough set model is a powerful tool for processing complete information systems. But it cannot handle incomplete information systems ,so the research of incomplete information is one of the research fields in rough set. The paper improves the tolerance relation model proposed by M. Kryszkiewcz to make it more flexible. Then significance of attributes theory is applied to propose a algorism of attributes reduction based on the improved model. Then an example is given to verify the algorism. At last, we give a comparison between improved algorism and traditional algorisms.