扩展粗糙集模型中依赖关系及应用

Application of dependency relation in expanded rough set models

对现有的几种扩展粗糙集模型进行分析比较之后,分别给出了这几种模型中对象依赖和知识依赖的定义,并且给出了对象部分依赖和知识部分依赖的定义及其度量.根据这些定义,提出了分别以对象依赖为依据进行聚类和以知识依赖为依据进行约简的算法.最后将这种以依赖作为依据的聚类和约简方法应用于不完备数据表和不完备决策表两个实例,在这些扩展粗糙集模型中都取得了较好的效果,从而验证了该方法的正确性和有效性.

After analyzing and comparing several existing expanded rough set models,the definitions of object dependency and knowledge dependency in the above models are provided.Furthermore,the partial object dependency′s definition,the partial attribute dependency′s definition and their dependency degrees are also presented.The uniform object reduction and knowledge reduction algorithms are given according to the dependency definitions in the above-mentioned models.Finally,the algorithms are applied to an incomplete data table and an incomplete decision-making table.Fairly good results obtained in these different expanded rough set models verify the correctness and validity of the method.