Large-Scale Data Envelopment Analysis Models and Related Applications

摘要: Abstract : This dissertation presents several advances in data envelopment analysis (DEA), a method for assessing the efficiency of decision units through identification empirical best-practice frontiers. First, new hierarchical decomposition approach solving large-scale problems is described with results computational testing both serial and parallel environments, that dramatically reduces solution time realistic DEA applications. Second, set models stratifying ranking provides important newer insights into relationships among than what was possible traditional frontier analysis. Because intensive requirements these models, their practicality builds on effectiveness process. Finally, means robustness decision-unit's given which spans all current assists managers evaluation process organizational improvement options. It expected will permit practitioners researchers to be more expansive ambitious use this class hopefully encourage even exciting applications DEA.