摘要： 对核属性的作用,以及求取属性核心的代价等进行了分析,并运用Rough集的理论给出了判定一个属性子集中是否包含属性核心的充要条件。根据这些研究结论对基于粗糙集的属性约简算法进行改进,并利用数据库查询语言实现了算法。 实验表明对于大数据集,该算法的效率大大高于一些基于主存的算法,且易于实现。

关键词: Rough集, 属性约简算法, 属性核心, 数据库技术

Abstract: The role of core attributes and the cost of calculating core attributes are studied . A criterion is proposed to determine whether all core attributes of a decision table are included in a subset of attributes. Based on these research results, an improved attribute reduction algorithm based on Rough set is presented . Experiments show that the algorithm is more efficient in large data sets and can be easily realized in advantage of database query language.

中图分类号: 

• TP311.13