基于多变量决策树的不完备信息系统规则提取研究  

Research on mining rules from incomplete information systems based on multivariable decision tree

作　　者：王妍妍 王艳宁[2] 刘永伟[3] 彭晓婷[4] 

机构地区：[1]燕山大学图书馆,河北秦皇岛066004 [2]燕山大学理学院,河北秦皇岛066004 [3]石家庄经济学院信息工程学院,河北石家庄050031 [4]燕山大学教务处,河北秦皇岛066004

出　　处：《燕山大学学报》2007年第3期221-224,共4页Journal of Yanshan University

摘　　要：不完备信息系统也有很多有用信息,可以得出支持决策的确定规则。基于决策树的完备信息系统规则提取方法很多,如比较有影响的ID3算法等,但是在不完备信息系统中应用决策树进行规则提取的方法还不是很多。本文提出用多变量决策树方法对不完备信息系统进行规则提取的方法,可以得出部分确定信息,方法简单易行。There is a lot of useful information in the incomplete information systems, certain rules can be acquired to make decision. In the complete information systems, there are many methods of mining rules based on the decision tree such as ID3 arithmetic,but there are not many methods of mining rules from incomplete systems based on the decision tree. In this paper, certain rules can be acquired by make multivariable decision tree in the incomplete information systems. This method is very easy to apply in the actual information systems.

关 键 词：不完备信息系统 相对核 多变量决策树 规则提取 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]