一种基于二进制区分矩阵的属性约简算法

An Attribute Reduction Algorithm Based on Binary Discernibility Matrix

摘要: 提出一种基于二进制区分矩阵的属性约简算法，将传统的区分矩阵以二进制形式表现出来，并且针对这个区分矩阵的特点，从两个不同的层次考察属性的重要性，引导求解过程趋于最优化．采用多个数据库进行试验，并与其它两种约简算法进行比较．实验结果表明，该算法是正确且高效的．

Abstract: An attribute reduction algorithm based on binary discernibility matrix is proposed,which represents the traditional discernibility matrix in a binary form,and according to the characteristics of this discernibility matrix,attribute significance is investigated at two different levels,which can lead the solutions to optimization.Experiments are carried out on various databases,and the presented algorithm is compared with other two reduction algorithms.The experiment results show that the proposed algorithm is valid and efficient.