A further generalization of midpoint convexity of multimaps towards common fixed point theorems and applications (Q526104)

Classical topological fixed point theorems for both single-valued and multivalued maps often require some kind of compactness or, in order to extend the applicability to the setting of reflexive infinite-dimensional Banach spaces, at least weak compactness. In this paper, the authors prove fixed point and common fixed point theorems in a more general framework which builds on the geometrical notion of (piecewise) midpoint linear operators introduced by \textit{S. Bolognini} and the authors [Fixed Point Theory 15, No. 1, 3--22 (2014; Zbl 1334.47051)], and extends previous work by \textit{M. Chermisi} and the second author [Fixed Point Theory Appl. 2008, Article ID 648591, 13 p. (2008; Zbl 1175.47053)]. They also introduce a still more general class of multivalued maps (labelled with the not very telling names ``strongly nice'' and ``weakly nice'') and study structural properties of their fixed point sets. As usual, such fixed point results have applications to, and are sometimes even equivalent to, best approximation results, complementarity problems, minimax theorems, and variational inequalities, either in the classical sense of Lax-Milgram and Stampacchia type, or in the modern sense of \textit{S. Park} [Nonlinear Anal., Theory Methods Appl., Ser. A, Theory Methods 73, No. 4, 1028--1042 (2010; Zbl 1214.47042)]. Hence this paper may be considered, apart from providing new results and applications, as a survey on such interconnections.