\(E\)-differentiable minimax programming under \(E\)-convexity (Q2241199)

This paper deals with a new class of minimax programming problems in which the involved functions are \(E\)-differentiable. For the original minimax programming problem with \(E\)-differentiable functions, its associated \(E\)-minimax programming problem is constructed in which the functions involved are differentiable in the usual sense. Both the so-called parametric and nonparametric necessary optimality conditions are derived for an optimal solution of the auxiliary \(E\)-minimax programming problem which, at the same time, are the so-called parametric and nonparametric necessary \(E\)-optimality conditions for an \(E\)-optimal solution of the original minimax programming problem. By using the parametric approach, sufficient optimality conditions are proved under \(E\)-convexity hypotheses. Subsequently, the necessary \(E\)-optimality conditions are applied to formulate the so-called \(E\)-dual problems for the considered \(E\)-differentiable minimax programming problems. Further, Mond-Weir and Wolfe duals are constructed for the \(E\)-minimax programming problem which, at the same time, are the so-called Mond-Weir and Wolfe \(E\)-dual problems for the original \(E\)-differentiable minimax programming problem. Then, several \(E\)-duality theorems are established between the \(E\)-differentiable minimax programming problem and its \(E\)-dual problems under appropriate \(E\)-convexity assumptions.