Abstract: In view of the hesitant fuzzy multi-attribute group decision making problems with the information of attribute weights which are completely unknown, a method based on the new reference point and prospect theory is proposed. The hesitant fuzzy decision making matrices have been turned into interval fuzzy decision making matrices, and then combined with the estimation of t-distribution, this paper investigates the expected value of the score function as the reference point. Based on the difference between the attribute value’ score function and the reference point, the value function and the prospect value matrix have been obtained. It applies the prospect value variance to calculate the attribute weights, and then the weighted prospect value of alternative is acquired on the base of the prospect theory. All the alternatives are ordered. A numerical example about the selection of cloud computing offerings is provided to illustrate the feasibility and effectiveness of the proposed approach.

Key words: multi-attribute group decision making, prospect theory, hesitant fuzzy set, expected value of score function

摘要： 针对属性指标值为犹豫模糊信息且属性权重完全未知的多属性群决策问题，提出一种基于新的决策参考点和前景理论的多属性群决策方法。将犹豫模糊决策矩阵转变为区间模糊决策矩阵，并结合t-分布估计方法，构建期望得分函数值，将其作为决策参考点；基于属性值得分函数与决策参考点之间的差异，确定价值函数，进而得到前景价值综合矩阵；利用同一属性下的前景价值方差计算属性的权重，并基于前景理论计算各方案的加权前景价值，进而对决策方案进行优劣排序；通过对云计算产品的选择实例验证提出的决策方法的可行性与有效性。

关键词: 多属性群决策, 前景理论, 犹豫模糊集, 期望得分函数值