粒度计算是粗糙集理论研究的一种强有力的工具。为了更好地研究不完备决策表中的知识获取方法,在基于粒度计算理论和极大相容块技术上,将不完备决策表分为确定的和不确定的两类,讨论了不完备决策表内在的粒度思想,并给出了两类不完备决策表在不同粒度下的相互转化关系。大量实验表明该方法是可行有效的。
Granular computing is emerging as a powerful tool for rough set theory. For researching the knowledge acquisition method better, in this paper based on granular computing theory and maximal consistent block, the incomplete decision table is classified as consistent and inconsistent, and the internal granular essence of the incomplete decision table is discussed, and the transform relations of the two kinds of the incomplete decision table is proposed under different granularity. Experiments show that this method is feasible,