Influence of deleting some of the inputs and outputs on efficiency status of units in DEA

One of the important issues in data envelopment analysis (DEA) is sensitivity analysis. This study discusses about deleting some of the inputs and outputs and investigates the influence of it on efficiency status of Decision Making Units (DMUs). To this end some models are presented for recognizing this influence on efficient DMUs. Model 2 (Model 3) in section 3 investigates the influence of deleting i(th) input (r(th) output) on an efficient DMU. Thereafter these models are improved for deleting multiple inputs and outputs. Furthermore, a model is presented for recognizing the maximum number of inputs and (or) outputs from among specified inputs and outputs which can be deleted, whereas an efficient DMU preserves its efficiency. Finally, the presented models are utilized for a set of DMUs and the results are reported.