a new inverse dea approach on classifying inputs and outputs: application in banking

In the original data envelopment analysis (dea) models, we assume from the beginning that the data has a specific role as input or output. but in many cases, we face a situation where some data can play the role of input or output, and based on maximizing the efficiency of the dmu under evaluation, they are selected as input or output in the model. in this paper, we present a new inverse dea model to classify these inputs and outputs. in the new model, we determine the new level of inputs and outputs and flexible measures by choosing the target efficiency for the decision-making units (dmus). in this regard, the new model may choose flexible measures as input or output, but the main goal is to reach the target efficiency level. in the following, we will explain the presented approach with a simple numerical example and show the application of the presented models in the banking industry. we also present the results of the models.