Convexlike alternative theorems and mathematical programming

From MaRDI portal





The author extends a theorem of the alternative for convexlike functions, due to \textit{M. Hayashi} and \textit{H. Komiya} [J. Optimization Theory Appl. 38, 179-189 (1982; Zbl 0471.49033)], to a larger class of functions, which are called \(\rho\)-convexlike. This is used to derive a saddlepoint theorem and a Lagrangian duality theorem for inequality constrained minimization problems defined by \(\rho\)-convexlike functions. For nondifferentiable problems, under a \(\rho\)-convexlikeness assumption on the directional derivatives of the functions, local necessary conditions of Fritz John and Kuhn Tucker types are presented. Another alternative theorem of Motzkin type for convexlike functions taking values in finite dimensional spaces, which extends a theorem of \textit{F. Giannessi} [ibid. 42, 331-365 (1984; Zbl 0504.49012)] for the convex case, is also given. This is applied to obtain new Fritz John and Kuhn Tucker type necessary optimality conditions for nondifferentiable problems, the directional derivatives of whose functions satisfy a convexlikeness condition.




Cited in
(58)








This page was built for publication: Convexlike alternative theorems and mathematical programming

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3705226)