论文部分内容阅读
广义DEA方法是相对于样本单元对决策单元进行评价,如何选择样本单元对于评价的结果有一定的影响.有时可以选择决策单元集的子集作为样本单元集,这样更容易被决策者所接受.通过构造符合一定条件的虚拟决策单元与所有决策单元一起采取最短距离法进行聚类分析,用包含虚拟决策单元的类作为样本单元集,可以使评价结果与传统DEA方法相比能够进一步体现出相对效率之间的差异.
The generalized DEA method evaluates the decision-making unit relative to the sample unit, and how to choose the sample unit has a certain influence on the result of the evaluation, and sometimes the subset of the decision-making unit set can be selected as the sample unit set, which is more easily accepted by decision makers. By constructing the virtual decision-making unit that meets certain conditions and all the decision-making units to carry out cluster analysis by using the shortest-distance method and using the class that contains the virtual decision-making unit as the sample unit set, the evaluation result can be further compared with the traditional DEA method to reflect the relative The difference between efficiency.