Geometric properties of Banach spaces and nonlinear iterations (Q2518094)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Geometric properties of Banach spaces and nonlinear iterations
scientific article

    Statements

    Geometric properties of Banach spaces and nonlinear iterations (English)
    0 references
    13 January 2009
    0 references
    This monograph gives an introduction to and overview of the author's extensive work on fixed point iterations. It consists of three parts. Part~1 (Chapters 1 to 5) is dedicated to the geometric properties of Banach spaces, namely, convexity, smoothness and the duality map. In Part~2 (Chapters 6 to 14), the author gives many results about fixed points of different classes of mappings. He focuses the main attention on the iterative processes that converge to a fixed point. The celebrated Banach contraction theorem assures that the Picard iteration formula \[ x_0 \in K, \quad x_{n+1} = T x_n \quad (n \geq 0) \] provides a sequence which converges to the unique fixed point, if \(K\) is a complete metric space and \(T: K \to K\) is contractive. If \(T\) is nonexpansive, other iteration processes are considered. Under certain conditions, the Mann iteration formula \[ x_0 \in K, \quad x_{n+1} = (1 - \alpha_n) x_n + \alpha_n T x_n \quad (n \geq 0), \] where \((\alpha_n)_{n \geq 0} \subset ]0,1[\), \(\lim_{n \rightarrow \infty} \alpha_n = 0\) and \(\sum_{n \geq 0} \alpha_n = \infty\), provides a sequence convergent to a fixed point. Throughout the book, the author considers different classes of operators: contractive, nonexpansive, quasi-nonexpansive, asymptotically regular, uniformly asymptotically regular, etc. In Part~3 (Chapters 15 to 22), common fixed points for (finite, countable) families of mappings are studied. The final Chapter~23 presents a lot of the same results on set-valued mappings, plus some general comments, examples and open questions. Each chapter contains a section of exercises and another of historical remarks. The book ends with 561 references, of which 100 are of the author, and a short index. It contains some minor mistakes: for example, the Goldstine theorem is called Goldstein theorem. Almost all the theorems of this monograph are due to the author and his collaborators.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    fixed point
    0 references
    0 references
    0 references