基于粗集的不完备信息系统的粗糙分类和属性约简

针对不完备信息系统(含有缺省数据或不精确数据),研究它的粗糙分类;并基于相容关系,将分布约简、最大分布约简、分配约简、近似约简引入不完备信息系统;且给出了最大分布约简的一种启发式算法:条件信息量约简算法。经实验检验,该算法是有效的。

Rough Set-based Classification and Attribute Reduction Under Incomplete Information Systems

Rough set theory is a relatively new soft computing tool, it can effectively analyze and process information systems. Traditionally, the information system is assumed to be perfect, i.e attribute values are not missing and supposed to be precise. However, imperfect information system is ubiquitous. In this paper, we investigate rough set and rough classification of imperfect information system, and several reduction methods like distribution reduction,maximum distribution reduction,assignment reduction and approximate reduction. Information quantity and conditional information quantity are defined to express indispensable attributes under incomplete information systems. Based on conditional information quantity, a heuristic algorithm for maximum distribution reduction is presented.Finally, the experimental result shows this algorithm can find its maximum distribution reduction for incomplete information system.