Approximate fixed points for discontinuous set-valued mappings (Q1298715)

The authors introduce the notion of measure of discontinuity and establish an existence theorem of approximate fixed point for discontinuous set-valued maps. Definition 1. Let \(F:X\to 2^X\) be a set-valued map, and \[ \delta_x(F)=\limsup_{r\downarrow 0}\sup_{z\in X\cap B(x,r)} \sup_{v\in F(z)}\inf_{y\in F(x)} \| v-y\|. \] \(\delta_x\) is called the measure of discontinuity of \(F\) at \(x\). The measure of discontinuity of \(F\) on \(X\) is defined to be \(\delta(F)=\sup_{x\in X}\delta_x(F)\). Definition 2. A point \(x\in X\) is said to be an approximate fixed point of a set-valued map \(F\) if there exists a point \(y\in F(x)\) such that \(\| x-y\| \leq\delta (F)\). The main theorem of this paper is Theorem 7: Let \(X\) be a nonempty compact convex subset of \(\mathbb{R}^n\), and let \(F\) be a set-valued map with closed convex values. Then there exists a point \(x\in X\) such that \(\min_{y\in F(x)} \| x-y\| \leq\delta(F)\). In particular, there exists a point \(y\in F(x)\) such that \(\| x-y\| <\delta(F)\), and so \(x\) is an approximate fixed point of \(F\). This theorem is a generalization of Kakutani's theorem in the sense of approximate fixed points to the large class of discontinuous set-valued maps. As a matter of fact, Theorem 7 reduces to Kakutani's theorem when \(\delta(F)=0\).