Approaches to developing combined DEA methods for analyzing comparative efficiency of complex systems

This article analyzes approaches to developing combined Data Envelopment Analysis (DEA) methods for assessing the comparative efficiency of complex systems. The paper examines the main problems of traditional DEA methods, such as the limitations of relative evaluation, the complexity of aggregating assessments, nonuniqueness of weights, and the lack of Pareto optimality. Various approaches to overcoming these limitations are proposed, including DEA modification with artificial boundaries, the use of secondary goal models, combination with other analysis methods, application of dynamic DEA, modification for working with negative data, and the use of nonlinear models. The article also discusses the advantages of combined approaches and outlines promising directions for further research in DEA methodology.