关键词: 多属性群决策, 区间直觉模糊集, 决策者风险偏好, 不完全属性权重信息, 交叉熵

Abstract: This paper proposes a new Atanassov’s interval-valued intuitionistic fuzzy values(AIVIFVs)multiple attribute group decision making(MAGDM)method considering the risk preference of decision makers with incomplete attribute weight information. We determine the weight of each decision maker(DM)with regard to each attribute by taking into account the similarity degree and proximity degree simultaneously. Next, given the fact that the risk preference of decision makers can have an effect on decision making results and prevent an AIVIF matrix from being weighted for lots of times, we transform the AIVIFV collective decision matrix into an interval value matrix by the risk preference coefficient of DMs. In addition, in order to determine the attribute weights objectively, we establish a interval-valued intuitionistic fuzzy values programming model based on cross entropy. Furthermore, we rank all the alternatives by the comprehensive interval values of alternatives. At last, to illustrate the detailed process of the new method proposed in this paper, we take a research and development project selection problem for example. And by comparison analysis, it can be proved that the proposed method in this paper is superior to the existing methods.

Key words: multiple attribute group decision-making, atanassov’s interval-valued Intuitionistic fuzzy set, risk preference, incomplete attribute weight information, cross entropy