区间互补判断矩阵一致性水平修正的离散化方法

Discretization method to improve the consistency level for interval reciprocal preference relation

为提高决策信息的可靠性,提出了一种改进区间互补判断矩阵个体一致性水平的离散调整方法.首先,针对互补判断矩阵给出一种启发式迭代算法用来识别、修改最不一致的偏好元素,改进其一致性水平.然后,基于最差一致性指标和最优一致性指标,把区间互补判断矩阵的一致性水平改进方法转化为两类互补判断矩阵的一致性水平改进方法.修正后的偏好取值属于最初的离散标度集,更易被决策者所接受.最后,通过与已有方法的对比分析,验证了所提出方法的有效性和适用性.

In order to enhance the credibility of decision information,this paper proposes a discrete adjustment method for improving the consistency level of interval reciprocal preference relation.Firstly,a heuristic iterative algorithm is given to identify and modify the most inconsistent judgments for reciprocal preference relation,improving its individual consistency level.Secondly,based on the worst consistency index and the best consistency index,the consistency level improvement method for interval reciprocal preference relation is reduced to the method for improving the consistency levels of two types of reciprocal preference relations.The revised preferences belong to the original discrete evaluation scale,which are more acceptable by the decision makers.Finally,the validity and applicability of the proposed approaches are verified by comparative analysis with the existing methods.