Surrogate Constraint Duality in Mathematical Programming

摘要: This paper presents a unified development of surrogate duality theory that is applicable to problems in which Lagrangean gaps limit the usefulness standard approaches. A dual created by generating single constraint replace original problem constraints, rather than absorbing these constraints into objective function as Lagrangean. We give necessary and sufficient conditions for optimality both with without imposition complementary slackness, also consider related “overestimating” may be used strategy bracket optimal value primal. The invite direct comparison those duality, demonstrating not only approach yields smaller (as first observed Greenberg Pierskalla), but giving precise characterization manner extent this occurs. Concepts parametric relative subgradients, par...