Improved Imprecise DEA Model with AHP Preference Cone

In this paper, we proposed an imprecise DEA (data envelopment analysis) model improved by analytical hierarchy process (AHP) preference cone. On the basis of traditional DEA model, we extended inputs and outputs of decision making units to interval issue, which formed imprecise DEA model. To reflect the preference of decision makers, AHP method was utilized to build the preference cone. It restricts the weights of imprecise DEA model. The hybrid model can provide us both objective and subjective solution. Possibility degree formula was applied to rank the interval efficiency scores. Finally, a numerical example illustrates the application of the improved model. The results show its availability and efficiency.