传统的基于数据包络分析的环境效率评价常常忽略投入、期望产出和非期望产出为离散型变量的可能性,从而无法精确处理现实中存在的具有整数特征的指标数据,一直以来影响着评价结果的精确性.本文以混合导向的整数DEA为基础,对超效率SBM模型进行拓展,构建了一种能够同时兼顾整数约束和非期望产出指标数据的超效率SBM模型,并在实例中采用搜索算法,对该模型进行求解并通过Bootstrap方法对结果进行修正.将每一个最优整数点与被评价决策单元之间的距离进行比较,可以找出最短距离所对应的最优整数点,即为其帕累托改进方向.该模型将为环境效率评价提供了适用范围更广泛的测度方法.
Traditional environmental efficiency evaluations based on DEA methods generally neglect that the inputs, desirable outputs and undesirable outputs are possibly discrete variables. Therefore, they cannot pre- cisely address the integer practical variables, which consequently affect the validity of the evaluations. This paper extends the super efficiency SBM model based on mixed-objective integer DEA and builds a new super- efficiency SBM model which considers both integer constraints and undesirable output variables. Searching al- gorithm is applied to solve the new model and Bootstrap method is further used to amend the results in an ap- plication. By measuring the distance between each optimal integer point and the evaluated decision making u- nit, the shortest one, i. e. , Pareto improvement direction, can be found. This model provides a more applica- ble tool for environmental efficiency evaluation.