基于DEA方法的额外资源分配算法

Arithmetic of extra resource allocation based on DEA method

研究基于DEA(数据包络分析)方法的资源分配问题.求解该类问题的难点在于:对于参与分配的决策单元来说,其分得的额外资源数不仅依赖于该决策单元的相对效率值,还依赖于它的规模大小,因此分配权重难以确定.根据输入、输出指标数的分布情况,建立了3种基本模型,分别是单输入单输出、多输入单输出、单输入多输出模型,给出了这3种模型的资源分配权重的确定算法.最后给出实例,并对给出的分配原则的合理性进行了分析讨论.

In this paper, we study the extra resource allocation problem based on DEA （Data Envelopment Analysis） method. The difficulty of this kind of allocation problem is that for every DMU （Decision Making Unit） which is involved in the allocation, the amount of the extra resource assigned to it depends on not only its relative efficiency but also its scale, so it is hard to choose allocation weights for every DMU. This paper gives the resource allocation algorithm according to three base models： single input and single output, single input and multi-output, multi-input and single output model in the case of input index and output index. Examples are presented, and the rationality of the allocation formula is also analyzed and discussed.