基于属性间交互作用指标的多属性决策方法

Approach to Multi-attribute Decision Making Based on Interaction Index among Attributes

多属性决策问题的各属性之间通常存在关联,而这种关联对决策的影响不可以忽略。基于属性间的交互作用指标,本文提出了一种新的评价函数,并研究了该评价函数的性质。首先,本文详细介绍了合作对策中几个常用的交互作用指标的概念。然后,提出了基于属性间交互作用指标的交互加权评价函数,简称为IWE函数。同时,本文给出了IWE函数的相关性质,并讨论了IWE函数与Choquet积分算子的联系与区别。最后,以算例验证本文所提方法在多属性决策中的合理性与有效性。

There is a correlation between the attributes which cannot be ignored in multi-attribute decision making problems. This paper proposes a new series of weighted evaluation functions based on interaction index among attributes in multi-attribute decision making. Some properties of the proposed evaluation function are also discussed. First, we introduce several frequently-used concepts of interaction index in cooperative games. Then, we propose a series of interaction weighted evaluation （IWE）functions based on referred interaction index. Some properties of these IWE functions are given. Meanwhile, the relationship and difference between IWE functions and Choquet integral are discussed. Lastly, a numerical example is used to illustrate the feasibility and validity of the proposed method.