基于可辨识矩阵的动态属性约简算法

The arithmetic of dynamic reduction based on differentiable matrix

可辨识矩阵是粗集理论中可以标识决策属性和条件属性之间关系的矩阵。根据数据库的动态建立特性,提出了一种基于粗集理论中的可辨识矩阵的动态属性约简算法,利用数据的动态性和可辨识矩阵的特性对已有的属性约简进行修正,并通过实例证明了这种算法的有效性,该算法极大地节省了计算量,提高了算法的执行效率。

Difiierentiable Matrix is a kind of matrix which may mark the relation ship between decision attributes and condition attributes in rough sets theory. According to the feature of buliding database dynamically, this paper proposes a kind of arithmetic of dynamic eduction based on differentiable matrix in rough - set theory, using the dynamic of data and the characteristic of differentiable matrix to modify the old attribute reduction and prove the validity of the arithmetic by examples. So the computation time is greatly saved and the efficiency of the arithmetic is greatly improved.