Iterative methods for equilibrium and fixed point problems for nonexpansive semigroups in Hilbert spaces (Q604247)

From MaRDI portal
Revision as of 07:23, 12 February 2024 by RedirectionBot (talk | contribs) (‎Changed an Item)
scientific article
Language Label Description Also known as
English
Iterative methods for equilibrium and fixed point problems for nonexpansive semigroups in Hilbert spaces
scientific article

    Statements

    Iterative methods for equilibrium and fixed point problems for nonexpansive semigroups in Hilbert spaces (English)
    0 references
    0 references
    0 references
    0 references
    10 November 2010
    0 references
    A continuous semigroup \(T(s)\), \(s\geq0\), in a Hilbert space \(H\) is called nonexpansive if \(\|T(s)x-T(s)y\|\leq\|x-y\|\) for arbitrary \(x,y\) in \(H\). For a bifunction \(G:H\times S\rightarrow\mathbb{R}\), \(EP(G)\) denotes the set of equilibrium points, i.e., all \(x\) such that \(G(x,y)\geq0\) for all \(y\in H\). The present paper studies iterative methods of finding solutions to a variational problem among equilibrium points, which are fixed points of \(T\). The authors present two iterative constructions, which they call implicit and explicit. In the implicit iterative construction, a pair of continuous sequences \(x_t,\) \( u_t\in H\) is defined as solution to \[ G(u_t, y) +\frac{1}{r_t}\langle y-u_t, u_t-x_t\rangle\geq 1, \quad \forall y\in H \] and \[ x_t = t\gamma f(x_t)+(I-tA)\frac1{\lambda_t}\int_0^{\lambda_t} T(s)u_t\,ds. \] Here, \(0<t<1\) and \(A\) is a strongly positive bounded linear operator. The explicit iterations are defined by the system \[ x_{n+1} = \alpha_n\gamma f(x_n)+(I-\alpha_n A)\frac1{s_n}\int_0^{s_n}t(s)u_n\, ds \] and \[ G(u_n, y)+\frac1{r_n}\langle y-u_n, u_n-x_n\rangle \geq 0, \quad \forall y\in H. \] It is demonstrated that, with an appropriate definition of the coefficients \(r_t, \lambda_t, \alpha_n, s_n\), the implicit sequences \(x_t, u_t\), and the the explicit sequences \(x_n\), \(u_n\), both converge strongly to the unique solution to variational inequality \[ \langle (\gamma f-A)z, p-z \rangle \leq0 \] on the set of fixed points of the semigroup \(T\) belonging to \(EP(G)\). No examples illustrate the theoretical results.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    equilibrium problem
    0 references
    fixed points
    0 references
    semigroup of nonexpansive mappings
    0 references
    variational inequalities
    0 references
    iterative algorithms
    0 references
    strong convergence
    0 references