Akademik Veri Yönetim Sistemi
ANADOLU 18TH INTERNATIONAL CONFERENCE ON APPLIED SCIENCES, Mardin, Türkiye, 19 - 21 Aralık 2025, ss.359-360, (Özet Bildiri)
Data Envelopment Analysis (DEA) is a non-parametric method frequently favored for efficiency measurement, utilized to assess the efficiency of DMU’S (decision making units) that yield the same outputs with the same inputs. DEA models are linear programming (LP) problems in which the objective function is minimized or maximized subject to certain constraints. Although DEA has a very practical use due to its non-parametric structure, the main requirement of the method is that the inputs and outputs must always be positive. In DEA, negative inputs and outputs are called undesirable input-output. The undesirable input-output situation is not limited only to being negative. According to DEA, either inputs must be reduced or outputs must be increased (or both). However, there are situations where the output produced by a system should not be increased. For example, in a recycling process, waste is considered an input for this system. From the company's perspective, the more waste, the better, but from an environmental perspective, waste should not be a product that should be increased. In this study, hypothetical examples are provided regarding the methods presented in the literature for dealing with undesirable input-output situations in DEA (the data in the examples are fictional). The aim of this study is to demonstrate that various solution methods are possible, rather than abandoning the use of DEA in cases such as the presence of negative inputs and outputs, or where the placement of some inputs and outputs does not seem logically appropriate but is necessited by the production process.