Abstract: The DEA model with entropy weight constraint（E-DEA model） is used to evaluate the agricultural input and output efficiency of 13 prefecture-level cities in Jiangsu Province in 2016. The relative efficiency of each decision-making unit is determined to be more consistent with the actual situation. At the same time, the ineffective decision-making unit is optimized by the inverse E-DEA model. The results show that the E-DEA model can distinguish the multiple decision-making units more effectively; the agricultural production efficiency of the 7 cities including Changzhou, Nantong, Xuzhou, Lianyungang, Huaian, Yancheng and Suqian is lower; the main reasons for the low agricultural production efficiency are the redundant input of agricultural machinery and the excessive use of chemical fertilizer.

Key words: E-DEA model, inverse E-DEA model, agriculture production efficiency, Jiangsu province