Evaluation of rejected cases in an acceptance system with data envelopment analysis and goal programming

Some DEA models have been proposed for acceptance or rejection based on a set of cases that have been previously classified. Also, a modified DEA-type linear programming model has been proposed to determine whether a new case must be accepted or rejected, depending on its location on, above, or below the sample frontier. However, these models assume that all attributes that characterize a case are discretionary. This paper extends these results by proposing a model that includes discretionary and non-discretionary attributes (inputs), and more important, a goal program which resembles a modified additive model to determine which characteristics must be changed (and by how much) in order to accept an initially rejected case. A real application is provided to illustrate the potential of the proposed ideas.