基于案例比较信息的多准则群决策分类方法

提出了一个基于案例比较信息的多准则群决策分类方法.该方法使用级别高于关系作为决策者的偏好模型.它要求决策者根据自己的偏好预先给出部分方案之间的比较信息.基于这些比较信息和决策者的重要度确定了具有最大可信度的群体一致案例比较信息集，并利用该信息集构建一个混合整数规划模型来计算方案的分类区间，从而引导决策者给出部分方案的精确分类信息.在此基础上，确定一个代表函数来获得其余方案的精确分类.最后通过一个实例说明了方法的可行性和合理性.

Multiple criteria sorting method based on pairwise comparisons of alternatives in the group decision making context

In this paper, a multiple criteria sorting method based on pairwise comparisons of alternatives in the group decision making context is proposed. The method uses the outranking relation as decision makers' preference model. It requires decision makers to provide pairwise comparisons of some alternatives. It chooses a consistent pairwise comparison set with the maximal total reliability degrees based on these pairwise comparisons and the importance degrees of decision makers. A mixed integer linear programming model is built to calculate the range of classes to which the alternative can be assigned so as to guide decision makers to specify the fittest class of some alternatives. Then a representative function is selected to determine the fittest class of the remaining alternatives. At last, a sorting example of venture investments is provided to demonstrate the feasibility and rationality of the proposed method.