基于粗糙集的不完备信息系统增量式属性约简

属性约简是粗糙集理论在模式识别中一项重要的应用，传统的属性约简算法只适合处理静态的信息系统，而处理不断动态更新的信息系统面临着巨大的挑战。对于不完备信息系统，提出一种增量式的属性约简算法。在不完备信息系统下引入粗糙集理论中关于正区域的概念，针对不完备信息系统中属性增加的情形，提出了基于正区域的增量式属性约简算法。实验结果表明了所提出的增量式属性约简算法比非增量式的算法具有更高的效率，同时比其他同类型的算法具有更高的优越性。

Incremental Attribute Reduction of Incomplete Information System Based on Rough Set Method

Attribute reduction is an important application of rough set theory in pattern recognition.The traditional attribute reduction algorithm is only suitable for dealing with static information system,however,handling constantly-dynamic updated information systems faces enormous challenges.For incomplete information system,an incremental attribute reduction algorithm is proposed.Firstly,the concept of positive region in rough set theory is introduced under incomplete information system.Then,an incremental attribute reduction algorithm based on positive region is proposed for the increase of attributes in incomplete information system.Finally,the experimental results show that the proposed incremental attribute reduction algorithm is more efficient than non-incremental algorithm,and has higher superiority than other similar algorithms.