Centralized additive model

While conventional data envelopment analysis (DEA) models set targetsseparately for each decision making unit (DMU), Lozano and Villa (2004)introduced the concept of "centralized" DEA models, which aim atoptimizing the combined resource consumption by all units in an organizationrather than considering the consumption by each unit, separately. In thesemodels, there is a centralized decision maker (DM) who supervises all DMUs. Themain aim is optimizing total input consumption and output production. In thispaper, firstly we present centralized output product model. Then we introduceparametric centralized additive model, which during one phase minimizes totalconsumption inputs and maximizes total output production simultaneously, in thedirection of optimization vector. Some numerical examples of the proposed modelsand their results are presented.