CAJ | 학술논문

　　信息系统中的属性约简是粗糙集知识发现的一个重要步骤。致力于研究一个信息系统中的特征选择、删除冗余属性。新的算法从属性重要性出发，采用迭代特征选择的标准，使得选择特征属性集不断缩小，获得信息系统的约简。通过实验证明该方法可行，有效。
　　신식계통중적속성약간시조조집지식발현적일개중요보취。치력우연구일개신식계통중적특정선택、산제용여속성。신적산법종속성중요성출발，채용질대특정선택적표준，사득선택특정속성집불단축소，획득신식계통적약간。통과실험증명해방법가행，유효。
Attribute reduction in information system is an important step during knowledge acquisition using Rough set. This paper focuses on the research of feature selection, deleting superfluous attributes in an information system. The new algorithm begins with the attribute significance, adopting iterative feature selection standard, making the selected feature attribute set get smaller, thus it acquires the reduction of information system. The experiment demonstrates that this method is feasible and effective.