Abstract: With respect to the multi-attribute decision making problem with correlated attributes, in which the decision maker’s reference dependence and loss aversion behaviors are considered, a novel C-TODIM decision making method is proposed. Based on the classical TODIM method, considering the reference dependence behavior of decision maker, the gain and loss values of each alternative relative to the other alternative concerning each attribute are calculated. Then, considering the loss aversion behavior of decision maker, the perceived dominance degree for each alternative over the other one is obtained by integrating the corresponding gain and loss values on multi-attribute. Furthermore, the total perceived dominance degree for each alternative can be computed to obtain the ranking result of alternatives. Finally, an example of venture investment is given to illustrate the feasibility and validity of the proposed method.

Key words: multi-attribute decision making, behavior, correlated attributes, C-TODIM decision making method, perceived dominance degree

摘要： 针对属性具有关联性的多属性决策问题,考虑到决策者具有参照依赖和损失规避行为,提出一种新的C-TODIM决策方法。依据经典TODIM决策方法,考虑决策者的参照依赖行为,计算每个方案相对于其它各个方案关于各属性的收益或损失值;再考虑到决策者的损失规避行为,集成属性关联情形下方案关于所有属性的收益或损失值,得到每个方案相对于其它各个方案的个体感知优势度;在此基础上,计算每个方案的总体感知优势度,并依据总体感知优势度的大小对方案进行排序。最后通过一个风险投资的算例验证该方法的可行性和有效性。

关键词: 多属性决策, 行为, 属性关联, C-TODIM决策方法, 感知优势度