The purpose of this paper is to propose a novel decision-making method based on fuzzy rough sets (FRSs) to deal with the uncertainty and imprecision existed in various multi-attribute decision-making (MADM) problems. In view of the effectiveness of fuzzy neighborhood operators in handling uncertain numerical data and the deficiencies of existing fuzzy neighborhood operators, we first define a reflexive fuzzy neighborhood operator in fuzzy information systems. Then, two types of FRS models are presented and their relationships are discussed. Whereafter, we use the tight FRS model to transform uncertain data into intuitionistic fuzzy data. A new MADM method is established under the intuitionistic fuzzy environment by using the idea of the PROMETHEE II and EDAS methods. Meanwhile, the intuitionistic fuzzy weights (IFWs) of attributes and the global intuitionistic fuzzy thresholds are introduced. Furthermore, a real-world example from the UCI database is utilized to expound the feasibility of the proposed method. At last, the validity and stability of the proposed method are demonstrated by comparative analysis and experimental analysis.

本文的目的是提出一种基于模糊粗糙集（FRS）的新颖决策方法，以解决各种多属性决策（MADM）问题中存在的不确定性和不精确性。鉴于模糊邻域算子在处理不确定数值数据中的有效性以及现有模糊邻域算子的不足，我们首先在模糊信息系统中定义了自反模糊邻域算子。然后，提出了两种类型的FRS模型并讨论了它们之间的关系。此后，我们使用紧密FRS模型将不确定数据转换为直觉模糊数据。利用PROMETHEE II和EDAS方法的思想，在直觉模糊环境下建立了一种新的MADM方法。同时，介绍了属性的直觉模糊权重（IFW）和全局直觉模糊阈值。此外，利用UCI数据库中的实际示例来阐述所提出方法的可行性。最后通过比较分析和实验分析证明了该方法的有效性和稳定性。