基于Rough集的决策规则提取

Decision Rules Extraction Based on Rough Set

通过分析分明矩阵方法进行属性约简时存在的一些弱点,本文提出了分明矩阵方法的简化,并将其应用于数据库中决策规则的提取。在求核集时,该方法不必生成分明矩阵这一环节,从而减少了时空上的不必要开销,提高了属性约简的效率,并通过一个实例详细地说明了具体提取过程,该实例说明了基于Rough集提取决策规则的方法简单易行。

By analyzing some weaknesses existing in the method of discriminable matrix during the process of reducing attributes, this paper proposes the predigestion of discriminable matrix which is applied into decision rules extraction in database, This method avoids creathag discriminable matrix in finding VCORE and reduces unnecessary spending in time and space, so the efficiency in reducing attributes can be improved. The procedure of decision rules extraction is illuminated using an example. It is simple for decision rules extraction based on Rough set by the example.