在VPRS模型中保持边界的属性约简方法研究

Research on Approximation Operator in Generalized Variable Precision Rough Set Model

讨论了变精度粗糙集模型中现有的属性约简方法,找出了β约简的不足;介绍了Inuiguchi提出的保持决策类下近似,上近似,边界和无法预言区的属性约简定义;说明了保持下近似的属性约简就是β下分布约简,保持上近似属性约简就是β上分布约简;提出了变粗度粗糙集模型中基于边界的属性约简方法,并从理论上证明了它的正确性;最后,给出了该种方法的实现算法。经实例证明,该方法操作简单,具有很高的应用价值。

Attribute reduction approaches at hand in variable precision rough set model are discussed, the shortages of reduct are found. Definitions of attribute reduction for preserving lower approximations, upper approximations, boundary regions and unpredictable region proposed by Inuiguchi are introduced. That attribute reduction approach of preserving lower approximations in β lower distribution reduction and approach of preserving upper approximations is β upper distribution reduction are demonstrated. Attribute reduction approach boundary regions-preserved in variable precision rough set model is proposed and its correctness is proved from theoretical. Finally, the algorthm of this approach is given The experiments show that this approach operates simply and it has better applied value.