Abstract: After analyzed the attribution reduction algorithm based on Rough Set that has arisen at present, we thought that the algorithm complexity of calculating importance degree of attribution could be further descended in existent attribution reduction algorithm. At the same time, a new and relatively reasonable formula measuring the importance of attribution was given, and the property of this formula was analyzed. Then we provided a quickly attribution reduction algorithm . At last, we explained the efficiency of the new algorithm with an example.

Key words: attribution reduction, algorithm complexity, Rough Set, Information system, attribution importance

摘要： 在分析目前已有的基于Rough Set的属性约简算法后, 认为计算属性的重要性的算法复杂度可以进一步降低, 同时给出了一个新的较为合理的度量属性重要性的计算公式, 并分析了该计算公式的性质, 然后给出了一个时间复杂度被大幅度降低的快速属性约简算法, 最后用一个实例说明了算法的有效性.

关键词: 属性约简, 算法复杂度, 粗糙集, 信息系统, 属性重要度