论文部分内容阅读
提出一种多属性群决策中同时考虑专家群体对属性主观赋权的偏好和决策者对决策重要性认识的偏好来确定属性权重的方法,能够兼容专家实数型、区间型和语言型等类型的属性权重赋值.首先建立标准属性重要差异矩阵以实现专家对属性赋权的优劣比较和差异规范,定义统一的决策者偏好映射对其进行调整;然后求解各矩阵的排序向量以量化属性的相对重要程度,并与专家权重聚合得到属性权重向量;最后给出了方法的具体步骤,并通过算例说明了该方法的具体应用.
This paper proposes a method to determine attribute weights by considering the preference of subgroups of experts in subjective weighting of attributes and the preferences of decision makers in understanding the importance of decision making in the multi-attribute group decision-making. It can be compatible with real-type, interval-type and linguistic types of experts Firstly, the important difference matrix of standard attributes is established in order to realize the experts’ comparison of the advantages and disadvantages of attributes and the difference norms, and to define the unified decision maker’s preference mapping to adjust them. Then, the ordering vectors of each matrix are calculated to quantify the relative attributes Degree of importance, and aggregated with expert weight to get the attribute weight vector. Finally, the concrete steps of the method are given, and the concrete application of the method is illustrated by an example.