Identifying the anchor points in DEA using sensitivity analysis in linear programming

Identification of the anchor points in FDH models

“…Furthermore, in Table 5, (j, p); λ jo < 1, µ jo p > 1 & (j, q); λ jo > 1, µ q jo < 1 . Bougnol and Dulá (2009) and its proof) and to some algorithms for convex case (see Mostafaee and Soleimani-damaneh 2014). This characterization and the related algorithms do not work in nonconvex case due to the lack of duality relations, lack of supporting hyperplanes at some boundary points and inequivalence of two above-mentioned definitions.…”

“…These points are far from the central part of the efficiency frontier. As can be seen from Theorem 1 ofMostafaee and Soleimani-damaneh (2014) and the results given byBougnol and Dulá (2009), the role of some input-output factors in efficiency situation of these units is not considerable. Anchor points were first used byThanassoulis and Allen ).…”

“…Thanassoulis et al, 2012;Mostafaee and Soleimani-damaneh 2014), the anchor points play a vital role in DEA theory and application. These points delineate the efficient frontier of the PPS from the free-disposability portion of the boundary (in convex DEA).…”

“…In a recently published paper, Mostafaee and Soleimani-damaneh (2014) presented an algorithm for identification of the anchor points utilizing sensitivity analysis techniques. See also Bougnol (2001) and Allen and Thanassoulis (2004) for more details about the notion and applications of the anchor points.…”

An algorithm for the anchor points of the PPS of the FRH models

“…Bougnol and Dulá [8] defined these points as the production possibilities which give the transition from the pareto-efficient frontier to the free-disposability portion of the PPS boundary and provided an approach to identify the anchor points of the variable returns to scale (VRS) PPS based on their geometrical properties. Mostafaee and Soleimani-damaneh [16] presented an algorithm for identifying of the anchor points by employing the sensitivity analysis techniques. The empirical applications led to somewhat surprising results in which that almost all extreme efficient units are in fact anchor points.…”