Resolving the infeasibility of the super-efficiency DEA based on DDF

Abstract

Chen et al. Omega 41:621-625, 2013 propose a DEA model which measures the super-efficiency with directional distance function. Their model avoids the infeasibility when the inputs have zero and it also prevents the negative projection. However, their model might be infeasible when there is zero in the outputs. Besides, their model relies on complex predetermined parameters. Recently, Lin and Chen (Lin and Chen Journal of the Operational Research Society 66:1506-1510, 2015) propose a new model which prevents infeasibility and negative projection. However, Lin and Chen's method introduces a constant which might dominate the input of the target DMU. To overcome the drawbacks of the models proposed in Chen et al. Omega 41:621-625, 2013 and Lin and Chen Journal of the Operational Research Society 66:1506-1510, 2015, we will develop a new DEA model to compute the super-efficiency, which prevents negative projection and the infeasibility when zero data occurs and works in the same way as the N-L model (Ray, Ray Journal of the Operational Research Society 59:788-797, 2008) when no abnormal input occurs.