基于差别矩阵的启发式属性约简算法及其应用  

Heuristic Attribute Reduction Algorithm Based on Discernibility Matrix and Its Application

作　　者：李三乐[1] 肖政宏[1] 

机构地区：[1]广东技术师范学院,广东广州510665

出　　处：《广东技术师范学院学报》2010年第6期11-14,共4页Journal of Guangdong Polytechnic Normal University

摘　　要：为了降低决策表属性约简算法的计算代价,利用属性重要度作为启发式运算因子,对基于SKowron差别矩阵的属性约简算法进行改进,并证明该方法的合理性.实例计算结果表明,在获得相同的结果下,该算法可以使计算量减少,提高计算效率.To reduce the computation of decision table’s attribute reduction algorithm,the attribute-impor-tance is used as a heuristic operation factor to improve attribute reduction algorithm based on SKowron discernibil-ity,It is proved that the algorithm is reasonable.The result shows that the presented algorithm reduces the compu-tation and improves the computational efficiency.

关 键 词：粗糙集 属性重要度 决策表 最佳属性约简 差别矩阵 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]