基于贴近度的异质多属性决策的重要性指数研究

Research on importance index of heterogeneous multi-attribute decision making based on closeness

针对评价信息为异质多属性决策问题,提出了一种基于相对贴进度的属性重要性指数的确定方法。首先,基于给定属性引起的模型值的平均变化,定义了一般离散模型的重要性指数,并证明其满足几个公理化。其中,模型没有限制性假设。然后,计算异质多属性决策问题中各决策方案的相对贴近度。基于贴近度构造值函数,进而确定各属性的重要性指数。最后通过实例计算与对比分析,验证了该方法的可行性和有效性。该方法有效地解决了异质属性类的重要性指数确定问题,且与传统方法相比更为精确,更适用于复杂决策系统。

In order to solve the problem of heterogeneous multi-attribute decision making,this paper proposed a method for determining the attribute importance index based on relative progress.First,based on the average change in model values caused by a given attribute,the method defined the importance index of a general discrete model and proved that it satisfied several axiomatizations.Among them,the model had no restrictive assumptions.Then,it calculated the relative closeness of each decision scheme in the heterogeneous multi-attribute decision problem.Next,it could determine the importance index of each attribute by the closeness construct value function.Finally,it verified an example and some contrastive analysis of the result to test the feasibility and effectiveness of the proposed approach.This method effectively solved the problem of determining the importance index of heterogeneous attribute classes,which were more accurate and more suitable for complex decision systems than traditional methods.