Vague决策系统属性约简算法研究

Research on Attribute Reduction Algorithm for Vague Decision Systems

属性约简是粗糙集理论最核心的研究内容之一,其中约简带有决策属性的决策系统的条件属性集可以达到简化信息系统目的.现有的粗糙集理论在处理属性值为精确值的决策系统方面已表现出强大的优势,但是在实际应用中要获得属性的精确值在有些情况下是困难的,面对这类信息系统,传统的粗糙集方法很难对其直接进行约简.为此,针对信息系统属性值为Vague概念的情况研究Vague决策系统的属性约简方法,通过借助包含度这一中间工具,设计一种在Vague决策系统中属性重要性的度量方法,并基于该方法设计了一种Vague决策表的属性约简算法,最后通过实例验证了该算法的有效性.

Attribute reduction is one of the important research issues in rough set theory. An information system is called decision systems which has decision attributes. The decision system can be simplified by reducing the condition attributes. The existing rough set theory had shown strong advantages in dealing with the decision systems which the attribute values are precise. However,to obtain the precise value of objects is difficult in the practical application. The traditional rough set theory cannot be directly to reduce the attributes of vague object information system. To address this issue,the tool of inclusion degrees is researched to vague decision system. The attribute significance is further designed for attribute reduction in vague decision system. On this basis,an attribute reduction algorithm based on the proposed measure is presented. Finally,an example is given to illustrate the effectiveness of the proposed algorithm.