Inexact orbits and boundedness of iterates of nonexpansive mappings (Q2822004)

From MaRDI portal





scientific article; zbMATH DE number 6629849
Language Label Description Also known as
default for all languages
No label defined
    English
    Inexact orbits and boundedness of iterates of nonexpansive mappings
    scientific article; zbMATH DE number 6629849

      Statements

      0 references
      0 references
      0 references
      26 September 2016
      0 references
      fixed point
      0 references
      holomorphic mapping
      0 references
      inexact orbit
      0 references
      iterates of nonexpansive mappings
      0 references
      Kobayashi distance
      0 references
      metric space
      0 references
      nonexpansive mapping
      0 references
      Inexact orbits and boundedness of iterates of nonexpansive mappings (English)
      0 references
      Right from the the first formulation and application of the celebrated Banach-Caccioppoli theorem, fixed point and other properties of nonexpansive mappings have been studied by many authors. There is a vast literature on various properties of such mappings; \textit{K. Goebel} and the third author of the present paper are among the leading specialists and have contributed with several monographs to the field [\textit{K. Goebel} and the third author, Uniform convexity, hyperbolic geometry, and nonexpansive mappings. New York, Basel: Marcel Dekker, Inc. (1984; Zbl 0537.46001)] and, more recently, the third author and \textit{D. Shoikhet} [Nonlinear semigroups, fixed points, and geometry of domains in Banach spaces. Hackensack, NJ: World Scientific (2005; Zbl 1089.46002)].NEWLINENEWLINEAdditional hypotheses which ensure the existence of a fixed point of a nonexpansive map in a normed or metric space are often quite subtle and are usually formulated in terms of the geometry of the invariant set. For example, if \(X\) is a Banach space, \(C\subseteq X\) is closed and convex, and \(T: C\to C\) is nonexpansive, it is known that the existence of a fixed point for \(T\) is closely related to the boundedness of the orbit sequence \((\overline x, T(\overline x), T^2(\overline x),\dots,T^n(\overline x),\dots)\) of some point \(\overline x\in C\).NEWLINENEWLINE Thus, a result of the following type is of interest, which is the main theorem of the paper. Given a metric space \((X,\rho)\) and a nonexpansive map \(T: X\to X\), assume that there exist \(\widetilde x\in X\) and \(R>0\) such that, for each \(m\in\mathbb{N}\), there are points \(x_{1,m},x_{2,m},\dots, x_{m+1,m}\) in the ball \(B(\widetilde x,R)\) such that NEWLINE\[NEWLINE\rho(T(x_{i,m}),x_{i+1,m})\leq{1\over m}\qquad (i= 1,2,\dots, m).NEWLINE\]NEWLINE Then the orbit sequence \((T^n(x))_n\) is bounded for each \(x\in X\). Some interesting corollaries of this are also given which in part generalize some results from a recent paper by the third author and \textit{A. J. Zaslavski} [Contemp. Math. 659, 273--281 (2016; Zbl 1388.46035)].
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references