Axiomatic approach to duality in optimization
Dinh The Luc, Johannes Jahn · Numerical Functional Analysis and Optimization · 1992
A new approach to duality in mathematical programming is developed for nonconvex vector problems with set-valued objectives. The approach is based on a weak duality axiom and a separation of nonconvex sets by means of monotonic functions. Strong duality results are obtained without convexity assumptions and without or under a rather weak constraint qualification. An application to nonconvex scalar programming is supplied.