Random fixed point theory for multivalued maps

DOI10.1080/07362999908809623MaRDI QIDQ4261535zbMATH OpenOpenAlexFDO

Publication date 18 December 2000

Published in Stochastic Analysis and Applications (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1080/07362999908809623

complements condensing map random fixed point theorems random operator

Set-valued operators (47H04) Contraction-type mappings, nonexpansive mappings, (A)-proper mappings, etc. (47H09) Fixed-point theorems (47H10) Random nonlinear operators (47H40) Random operators and equations (aspects of stochastic analysis) (60H25) Stochastic analysis (60H99)

Review text

Let \(E\) be a Banach space, \(C\) a closed convex subset of \(E\) and \(U\) an open set of \(C\). We denoted by \(\text{CK}(E)\) the family of all nonempty convex compact subsets of \(E\). We call \(N:\overline U\times [0,1]\to \text{CK}(C)\) a condensing map if \(\alpha(N(W))\leq \alpha(\pi W)\) for all bounded sets \(W\) of \(\overline U\times [0,1]\) and \(\alpha(N(A))< (\pi A)\) for all bounded nonprecompact subsets \(A\) of \(\overline U\times [0,1]\), where \(\pi:\overline U\times [0,1]\to\overline U\) is the natural projection and \(\alpha\) is the Kuratowski measure of noncompactness. \(K_{\partial U}(\overline U,C)\) denotes the set of all u.s.c. condensing maps \(F:\overline U\to \text{CK}(C)\) with \(F(\overline U)\) a subset of a bounded set in \(C\), and with \(F\) fixed point free on \(\partial U\) (i.e. \(x\notin Fx\) for all \(x\in\partial U\)). A mapping \(F\in K_{\partial U}(\overline U, C)\) is essential (in the Granas sense) if every \(G\in K_{\partial U}(\overline U, C)\) which agrees with \(F\) on \(\partial U\) has a fixed point in \(U\). Let \((\Omega,{\mathcal A})\) a measurable space. A mapping \(F:\Omega\times z\to 2^X\), where \((X,d)\) is a metric space and \(Z\subset X\) is called a random operator if for every \(x\in Z\), the map \(F(\cdot, x):\Omega\to X\) is measurable. In this paper new fixed point and random fixed point theorems for set-valued maps are proved. The main result of section 2 is Theorem 2.2: Let \(Q\) be a closed convex of a Hilbert space \(E\). Assume \(N:Q\times [0,1]\to \text{CK}(E)\) is a u.s.c. condensing map with \(N(Q\times [0,1])\) a subset of a bounded set in \(E\). Define \(r: E\to Q\) by \(r(x)= P_Q(x)\) (i.e., \(r\) is the nearest point projection on \(Q\)) and suppose \(N(r(\cdot),0)\in K_{\partial U}(\overline U, E)\) is essential for any open set \(U\) of \(E\) with \(Q\subset U\) and \(Q\cap \partial U=\emptyset\). In addition assume: if \((x_j,\lambda_j)^\infty_{j= 1}\) is a sequenc in \(\partial Q\times[0, 1]\) converging to \((x,\lambda)\) with \(x\in N(x,\lambda)\) and \(0\leq \lambda< 1\), then there exists \(j_0\in \{1,2,\dots\}\) with \(N(x_j, \lambda_j)\subset Q\) for each \(j\geq j_0\). Then \(N(x,1)\) has a fixed point in \(Q\). In section 3, the author presents a number of random fixed point results which extends and complements previously known results by \textit{K. Deimling}, \textit{G. S. Ladde} and \textit{V. Lakshmikantham} [Stochastic Anal. Appl. 3, 153-162 (1985; Zbl 0566.60058)]; \textit{S. Itoh} [J. Math. Anal. Appl. 67, 261-273 (1979; Zbl 0407.60069)]; \textit{T.-C. Lin} [Proc. Am. Math. Soc. 103, No.~4, 1129-1135 (1988; Zbl 0676.47041)]; \textit{D. O'Regan} [Proc. Am. Math. Soc. 126, No.~7, 1963-1971 (1998; Zbl 0894.47056) and Comput. Math. Appl. 35, No. 4, 27-34 (1998)]. In addition, the author extends recent results of \textit{K.-K. Tan} and \textit{X.-Z. Yuan} [J. Math. Anal. Appl. 185, No. 2, 378-390 (1994; Zbl 0856.47036)].

Recommendations

Cites work

Cited in

(15)

This page was built for publication: Random fixed point theory for multivalued maps

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4261535)