Thesis (M.Sc.) - Cairo University - Faculty of Economics and Political Science - Department of Statistics

Data Envelopment Analysis (DEA) is a data-oriented approach for evaluating the performance of a set of peer entities called Decision Making Units (DMUs) which convert multiple inputs to multiple outputs. DEA models rarely handle the problem of having multi-level hierarchically structured DMUs. In addition, most of these models are deterministic. This thesis has developed both generalized deterministic and stochastic multi-level DEA models. During model implementation, the issue of properly choosing the non-Archimedean infinitesimal has arisen. An objective technique for determining it has been proposed and tested. A reassessment of Egypt{u2019}s human development has been performed using the suggested deterministic multi-level DEA model. It has given more reasonable and consistent results than conventional Human Development Index (HDI). The suggested deterministic model has shown close results compared to the two-level model and significant superiority compared to the three-level model. The suggested stochastic model has been applied to a real dataset. It has given better results than conventional chance-constrained DEA model

Issued also as CD