基于数组的增量式属性约简研究

Research for incremental updating algorithm for attribute reduction based on array

为了在动态变化的决策表中,快速方便地获得新的属性约简,利用数组元素对差别矩阵中的属性组合进行计数,提出一种基于数组的增量式属性约简更新算法。当决策表动态变化时,根据数组元素的取值变化,获得差别矩阵中增加和减少的元素,依据这些元素更新原属性约简,快速得到新的最小属性约简。理论分析及实例验证了提出的算法是有效的。

Using array elements to count attribute combination in discernibility matrix,this paper introduced an incremental updating algorithm for attribute reduction based on array in order to obtain new attribute reduction quickly and conveniently in the dynamic decision table.When the decision table was in dynamic change,attribute combination increased and reduced could be obtained in discernibility matrix by the changes of the array element value.The old attribute reduction set could be updated effectively based on these attribute combination and the smallest attribute reduction set could be obtained quickly.The algorithm is proved to be efficient both by theoretical analysis and illustration of examples.