摘要
以同时具有丢失型和遗漏型未知属性值的不完备系统为研究对象,根据特征关系讨论了其中的知识约简问题.在不完备决策系统中,引入了近似分布约简的概念并给出了相应的判定定理与辨识矩阵,为从复杂的不完备系统中获取知识提供了新的理论基础与技术手段.
In this paper, the incomplete information system in which lost and ＂do not care＂ unknown values are coexisting is thoroughly investigated. Based on the characteristic relation, approach to knowledge reduction in the incomplete information system is studied. From the viewpoint of characteristic relation based rough set model, the approximate distribution reduction is introduced into the incomplete decision system. The judgment theorems and discernibility matrixes associated with these knowledge reductions are also obtained. The research is meaningful both in the theory and in applications for the acquisition of rules in the complex incomplete information systems.
出处
《江南大学学报（自然科学版）》
CAS
2007年第6期690-693,共4页
Joural of Jiangnan University (Natural Science Edition)
基金
国家自然科学基金项目(60472060
60572034)
国家自然科学基金重点项目(60632050)
江苏省自然科学基金项目(BK2006081)
关键词
不完备信息系统
特征关系
粗糙集
知识约简
incomplete information system
characteristic relation
rough set
knowledge reduction
作者简介
杨习贝（1980-），男（回族），计算机应用技术专业博士研究生．
通讯联系人：杨静字（1941～），男，河北秦皇岛人，教授，博士生导师．主要从事计算机视觉、模式识别、数字图像处理等研究．Email：yangjyr@mail．edu．cn
