基于差别矩阵属性约简算法的改进

Updating Algorithm for Attribute Reduction Based on Difference Matrix

基于差别矩阵的粗糙集属性约简改进算法,先判断差别矩阵核是否为空,再根据判断结果采取求差别矩阵、求核、属性约简等步骤,完成其相对约简。即通过构造决策表的差别矩阵核心程序求差别矩阵,在对属性进行约简前先对核进行判断并求核。最后根据输入输出进行属性约简。该方法能节约属性约简时间,提高算法效率。

The updating algorithm for attribute reduction in rough set theory based on difference matrix; at first, judging whether the core of the difference matrix is empty, then according to judging outcomes, seeking difference matrix and core, achieving attribute reduction steps, and completing its relative reduction. Seeking difference matrix by constructing the difference matrix core program of the decision-making table, judging core and seeking core before attribute reduction. Finally, doing attribute reduction according to the input and output. This method can save time of attribute reduction and improve the efficiency of algorithms.