摘要
从矩阵视角研究不完备信息系统的知识表示和属性约简.首先,引入关系矩阵,基于容差关系提出不完备决策信息系统协调性判定方法.其次,利用关系矩阵求正域,并提出属性协调集的矩阵判定定理,从而给出一种保持正域不变的约简的新方法.最后,基于上述框架,给出属性重要性度,进一步提出一种属性约简的启发式算法,并通过分析和实例证明该方法的有效性.
This paper studies the knowledge representation and attribute reduction of incomplete information systems from the perspective of matrix. Firstly, the relation matrix is introduced, and based on tolerance relation, a method for judging the consistent of incomplete decision information system is proposed. Secondly, the positive domain is obtained by using the relation matrix, and the matrix decision theorem of the attribute consistent set is proposed, and a new reduction method keeping the positive domain invariant is given. Finally, based on the above framework, an attribute importance significance is given, a heuristic algorithm for attribute reduction is proposed, and the effectiveness of the method is proved by analysis.
作者
陈应生
李进金
CHEN Yingsheng;LI Jinjin(School of Mathematical Sciences,Huaqiao University,Quanzhou 362021,China;School of Mathematics and Statistics,Minnan Normal University,Zhangzhou 363000,China)
出处
《华侨大学学报（自然科学版）》
CAS
北大核心
2019年第6期823-829,共7页
Journal of Huaqiao University(Natural Science)
基金
国家自然科学基金资助项目(11871259,61379021,11701258)
关键词
不完备决策信息系统
关系矩阵
容差关系
正域
incomplete decision information system
matrix of relation
tolerance relation
positive domain
