基于粗糙集的数据挖掘改进属性约简算法研究

Research on the improved algorithm for attribute reduction in rough set based on data mining

研究基于粗糙集的属性约简算法在数据挖掘规则提取阶段的应用。数据挖掘中对属性进行约简时,经常采用粗糙集,再按照规则进行提取。考察差别矩阵的定义和信息系统比较复杂且核属性元素所占比例较少的情况,改进基于差别矩阵的属性约简算法,利用差别矩阵的结构建立一种新的选择属性的依据。

Application of rule extraction algorithm of attribute reduction based on Rough Set Theory is studied in data mining.The data mining of attribute reduction is often based on rough set,and then it is done in accordance with the rules extraction.Based on the investigation to the relevant definition of discernibility matrix,the informa-tion system is more complex and nuclear property element accounts for a small proportion.The algorithm of attribute reduction based on discernibility matrix is improved by using the discernibility matrix structure to establish the basis for a new selection property.