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.
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
关键词
粗糙集
属性重要度
决策表
最佳属性约简
差别矩阵
rough set
attribute-importance
decision table
best attribute reduction
discernibility matrix
