Rough集之间的相似度量  

Measure of Similarity between Rough Sets

作　　者：徐久成 沈钧毅[1] 王国胤[2] 

机构地区：[1]西安交通大学电信学院软件所 [2]重庆邮电学院计算机科学与技术研究所,重庆400065

出　　处：《计算机科学》2003年第10期55-57,60,共4页Computer Science

基　　金：国家自然科学基金(69803014);国家科技部攀登--特别支持费;教育部骨干教师资助计划(GG-520-10617-1001);教育部留学回国人员科研启动基金;重庆市应用基础研究基金;信息产业部计算机网络与通信技术重点实验室研究基金

摘　　要：1引言 近年来由Z.Pawlak教授提出的经典Rough集理论在边界不确定信息处理和知识获取方面取得了很大成功.但遗憾的是,它是基于完备信息系统这样一个假设的,即所处理的信息表都是完备的,每个样本对象的所有属性值都是已知的.Applications of rough set theory in incomplete information systems are a key of putting rough set into real applications. In this paper, after analyzing some basic concepts of classical rough set theory and extended rough set theory, the measure of similarity is developed between two rough sets in the classical rough set theory based on indiscernibility relation and between two rough sets in the extended rough set theory based on limited tolerance relation. Then, some properties of these two methods for measuring similarity are developed respectively. At last, these two measure methods of rough set theory are compared.

关 键 词：ROUGH集 相似度量 信息处理 知识获取 智能系统 推理 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]