摘要
研究了信息系统(包括完备和不完备信息系统)基于信息量的属性的相关性及其在属性约简中的应用。引进了具有补集特征的信息量、联合信息量、条件信息量、交互信息量等概念,并给出这些概念的一些重要关系式。利用信息量定义属性间的相关系数,并利用相关系数来刻画一个信息系统是否是完备的这一特征,讨论了所引进的概念在信息系统上的知识发现中的一些应用,提出了一个基于属性相关系数的信息系统知识约简的一种新算法。实验结果表明了该算法的有效性。
Based on information quantity, the relativity between two attributes of information systems （including complete and incomplete information systems） and the relativity＇s application in the attribute reduction are studied. In the view of complement set, such concepts are introduced as information quantity, unite information quantity, conditional information quantity and mutual information quantity. Some important equations of these concepts are given. The definition of correlation coefficient based on information quantity is presented, based on correlation coefficient, some applications in knowledge discovery is discussed, and a new attribute reduction is presented. Finally, the experimental result shows that this algorithm is effective.
出处
《计算机工程与设计》
CSCD
北大核心
2012年第3期1192-1196,1217,共6页
Computer Engineering and Design
关键词
信息系统
信息量
相关系数
属性约简
算法
information systems information quantity correlation coefficient attribute reduction algorithm
