容差粗糙集的结构化近似

Structured Approximations in Tolerance Rough Set

将结构化技术用于容差粗糙集模型,提出结构化容差粗糙集模型,并借助φ算子对该模型进行刻画,研究其相关性质。由于结构化容差粗糙集的上、下近似是以容差类为元素的多重集,因而有效保地留对象间的结构信息。最后,对该模型的不精确性和不完备性的刻画进行讨论。

To some extent,the tolerance relation can overcome the problem that the indiscernable relation is too strict in the classical rough set theory,and the structuralization of the rough set can preserve the classified information of the data object.In this paper,the structuring technology is applied to the tolerance rough set model,and the structured tolerance rough set model is proposed.The proposed model is characterized by operator φ,and its related properties are also explored.Since the upper and lower approximations of structured tolerance rough set are multisets of tolerance classes,they can effectively retain the structural information of objects.Finally,the inaccuracy and the incompleteness of the proposed model are discussed.