A duality approach to minimax results for quasi-saddle functions in finite dimensions (Q1194857)

An extended-real function \(K(y,v)\) on a convex set \(A\times B\subseteq \mathbb{R}^ n\times \mathbb{R}^ m\) is termed quasi-saddle if it is quasiconvex in \(y\) and evenly concave in \(v\). The question is whether or under which conditions quasi-saddle functions have saddle values or, eventually, saddle points. The relevance of the question is lying in its intimate connection with duality theory. An elegant answer is provided in Theorem 6.