不完备信息决策规则提取的粗糙集方法  

Extracting Decision Rules from Incomplete Information Systems by Rough Set Approach

作　　者：梁美莲[1] 梁家荣[2] 范平[3] 李天志[4] 

机构地区：[1]广西大学数学与信息科学学院,广西南宁530004 [2]广西大学计算机与电子信息学院,广西南宁530004 [3]咸宁学院信息工程学院,湖北咸宁437000 [4]德州学院计算机系,山东德州253023

出　　处：《微计算机信息》2008年第33期245-246,238,共3页Control & Automation

基　　金：国家自然科学基金(60564001);广西大学科研基金(X061035)

摘　　要：粗糙集理论在不完备信息系统中应用的研究越来越受到学者们的关注,多个不完备信息系统粗糙集扩展模型先后建立起来。基于限制容差关系的扩展模型是其中一个较为灵活和符合数据挖掘需要的模型。本文基于限制容差关系,将论域分为两个部分,并针对两类不同的对象分别给出了约简的定义,在此基础上提出了决策规则提取和决策规则评估算法。Rough set theory being applied in incomplete information system has been increasingly concerned these years, and as a resuit, several extensions of rough set under incomplete information system have been established. Among them, the extension based on the τ limited tolerance relation is flexible and suitable for data mining. In this paper, the domain is divided into two parts, according to the T limited tolerance relation. Concepts of reduction are defined respectively for the two different objects, based on which novel algorithms of decision rules extracting and decision rules evaluating are proposed.

关 键 词：数据挖掘 粗糙集 不完备信息系统 容差关系 决策规则提取 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]