一种基于粗糙集的属性值约简方法  

ATTRIBUTE VALUE REDUCTION ALGORITHM BASED ON ROUGH SET THEORY

作　　者：陈晓云[1] 兰聪花[1,2] 

机构地区：[1]兰州大学信息科学与工程学院,甘肃兰州730000 [2]兰州工业高等专科学校电子信息工程系,甘肃兰州730050

出　　处：《计算机应用与软件》2010年第8期198-200,共3页Computer Applications and Software

摘　　要：利用矩阵的可操作性等优点,将决策表中的知识通过区分矩阵反映出来,对决策表的值约简操作转换为对矩阵的运算。提出基于区分矩阵的一些重要性质及在此基础上的属性值约简方法,最后通过实验和实例验证了该算法是有效可行的。The operability advantage of the matrix is utilised, the decision table knowledge is reflected with discernable matrix, the reduction operation on value of the decision table is converted into the operation on matrix. Some important properties based on discernable matrix are proposed together with the attribute value reduction method on this basis. Finally, the effective feasibility of the algorithm are verified in experiment and instance.

关 键 词：粗糙集 区分矩阵 属性值约简 属性重要度 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]