基于后悔理论的混合型多属性案例决策方法

本文针对决策者的后悔规避行为对案例决策的影响，提出了一种基于后悔理论的混合型多属性案例决策方法。首先，针对案例属性的相似度测算，将案例属性的范围拓展到定性数据、清晰数、区间数、语言变量和区间直觉模糊数五种类型；然后，在案例原有效用基础上考虑决策者的后悔-欣喜值来测算各案例的感知效用；进一步，通过案例相似度对感知效用加权求和得到备选方案的综合感知效用，并依据综合感知效用大小对备选方案进行排序；最后通过一个实例，验证了上述方法的可行性和有效性。

Case-Based Decision Analysis Method based on Regret Theory for Hybrid Multiple Attributes Decision Making

Although case-based decision theory is prevalent with solving decision issues, the main drawback is that decision makers' psychology has not taken into account. With the foundation and constant improvement of lifelong accountability system and responsibility for the investigations mechanism, it is nonnegligible to consider personal psychological factors. Therefore, how to integrate psychological and behavioral characteristics of decision-makers with case decisions becomes a concern for research question. Thus, in this paper, a multiple-attribute decision making method based on regret theory is proposed. In terms of regret theory, the regret-rejoice value is presented to represent the impact of psychological behavioral factors on the utility of decision-making, and a model with regret aversion, which integrates the cognitive limitations and psychological factors into the decision-making framework is built. At the same time, considering the complexity and uncertainty of reality, and the diversity of information, the case attributes are expanded to five types, including qualitative data, crisp data, interval data, linguistic variables and interval intuitionistic fuzzy data. Then methods for calculating similarity are presented by layering computing and exponential function. Specifically, first, similarity is calculated with an improved method proposed by the authors, so as to screen the approximation case set. Second, the objective utility of cases is calculated, the utility with regret theory is adjusted to get perceptible utility, then perceptible utility is integrated with similarity and the alternatives are ranked. Finally, taking the site selection of a PX project as an example and comparing with CBR and traditional CBDT, the result of this method is found relatively reliable, no matter in terms of wind direction, distance from water sources to cities, and the destructiveness of natural disasters. In addition, it is more in line with the actual decision process in context of the accountability mechanisms, reflecting the regret aversion behavior of decision makers. To sum up, the adaptability for case-based decision theory is expanded to complex multiple-attribute problems. Besides, the objective case is combined with decision maker's subjective feelings, and the effect of decision is taken maker's regret-aversion on final decision is taken into the model; finally, the feasibility and effectiveness of the proposed method are demonstrated by a site selection example.