信息系统中的局部邻域粗糙集及属性约简

Local Neighborhood Rough Sets and Attrib-ute Reduction in Information Systems

摘要: 信息粒度和近似方法是粗糙集理论中两个重要描述数据的方法，为了解决经典邻域粗糙集的计算效率低下和数据识别有用性不足的问题，有人提出了局部邻域粗糙集模型。在局部邻域粗糙集中，邻域半径的大小直接影响数据的有效性，因此邻域半径的取值为至关重要，但在局部邻域粗糙集中邻域半径的取值范围有点过大。为了进一步缩小邻域半径取值范围，本文借助邻域半径的取值，讨论了系统的协调性。首先给出了不协调信息系统中的局部邻域粗糙集，然后将属性约简不协调集的思想引入邻域，缩小了邻域半径的取值范围，为计算局部邻域粗糙集的邻域半径提供了有效工具，同时研究了相关的性质。

Abstract: The information granular and approximation method is two important methods to describe data in rough set theory. In order to work out the computational inefficiency and lack of cognitive data in the classical neighborhood rough set, then someone proposed the local neighborhood rough set model. For the local neighborhood rough set model, the value of neighborhood radius is very of great importance since its value has a close relation with the efficiency. However, the value range of neighborhood radius in the local neighborhood rough set is too large. Further, in order to reduce the value range of neighborhood radius, this paper studies the coordination of the system. Firstly, the local neighborhood decision rough sets of the uncoordinated information system are given, then the idea of the property reduction coordination set is added to the neighborhood, which can reduce the range of radius of the neighborhood, provide an effective tool for calculating neighborhood radius of rough set. The related properties are also studied.