区间数据包络分析的主客观求解

根据是否需要决策者的主观判断，将区间DEA的求解方法分为主观方法和客观方法．定义了一种反映决策者满意度的区间数序关系，基于此将区间DEA转化为一个确定型DEA，并求得给定满意度水平下的点效率值，此为主观方法．建立了求解每个DMU的最高效率值和最低效率值的确定型DEA模型，从而求得每个DMU的区间效率值，此为客观方法．分析了这两种方法在求解过程中所涉及到的数据一致性问题，即每一个DMU的区间投入或产出在确定型DEA中的所有出现要保持相同的投影点．

Subjective and objective method of solving interval DEA

According to whether the decision-maker's subjective judgment is needed, the method of solving interval DEA is classified into the subjective method and objective one. An order relation between interval numbers expressing the decision-maker's satisfaction degree is defined first. Based on this, an interval DEA is transformed into an accurate DEA and point efficiency scores are achieved at a given satisfaction degree. This is called as the subjective method. Then accurate DEA models of obtaining the highest and lowest efficiency score of each DMU are proposed. As a result, an interval efficiency score of each DMU is gotten. This is called as the objective method. The problem of data consistency existing in the both methods is analyzed. That is to say, the interval input or interval output of each DMU must have the same projection point throughout the accurate DEA.