摘要
在不完备信息系统中,利用容差关系提出的相似矩阵,给出了一种基于属性重要度的约简算法。在相似矩阵中,属性的重要度反映了该属性在区别对象时的能力,重要性越小的属性在相似矩阵中出现的次数越多,因为,此属性反映了对象的相似度较高。因此,利用属性的重要度来选择属性逐次加入约简集中,提出了一种新的算法。最后通过实验说明了该方法的正确性与有效性。
This paper proposes an attribute reduction algorithm based on attribute significance in the incomplete information system. The algorithm makes use of the concept of similar matrix via tolerance relationship. In the similar matrix, attribute significance reflects the ability of distinguishing between objects. The more frequent the appearance times are, the less importance the attribute is. The attribute reflects the higher similarity of objects. Then a new algorithm is presented which adds the attribute into the reduction set based on the attribute significance. Experiment results show that the algorithm is correct and effective.
出处
《长春工业大学学报》
CAS
2012年第4期456-459,共4页
Journal of Changchun University of Technology
关键词
不完备信息系统
相似关系
属性重要度
属性约简
incomplete information system
tolerance relationship
attribute significance
attribute reduction.
