Characterization of generalized FJ and KKT conditions in nonsmooth nonconvex optimization (Q2301197)

Fritz John (FJ) and Karush-Kuhn-Tucker (KKT) optimality conditions are investigated, without assuming a constraint qualification. This is done through the introduction of ``FJ points'' and ``KKT points'', as in [\textit{F. Flores-Bazán} and \textit{G. Mastroeni}, SIAM J. Optim. 25, No. 1, 647--676 (2015; Zbl 1398.90126)]. The innovation here is that, in contrast to the paper by Flores-Bazan and Mastroeni [loc. cit.], the functions are nonsmooth (locally Lipschitz), so the approach is in terms of the Clarke generalized gradient. An application to mathematical programs under a constraint defined by a variational inequality is also considered.