基于偏序集理论的数据包络分析方法,通过引进适当的偏序关系,挖掘出决策单元之间的特殊关系。然而,随着决策单元所选取的投入产出指标个数的增加,决策单元之间的偏序关系变得越来越少。对此,通过引进决策单元之间的距离和适当的样本决策单元,建立决策单元之间的特殊关系,最终生成决策单元之间的格论关系,并引进相关定理及其算法。最后通过仿真结果表明了所提出算法的有效性和实用性。
With the introduction of proper partial order relation in the data envelopment analysis method based on the partial ordered set theory, the special relationships between different decision making units are established. However, with the increase of the number of index of input and output data, the relationships between different decision making units decrease. With the introduction of distances between decision making units and proper sample decision making units, the relationships between different decision making units are established, the relations based on the lattice theory of decision making units are provided, and related theorems and algorithms are introduced. Finally, simulation results show the effectiveness and practicability of the proposed method.