Implicit iterative algorithms for asymptotically nonexpansive mappings in the intermediate sense and Lipschitz-continuous monotone mappings (Q848545): Difference between revisions

Suppose \(A\) to be a mapping of \(C\) into the real Hilbert space \(H\) where \(C\) a nonempty convex of subset \(H\) and let \(S:C\rightarrow C\). Denote by \(F(S)\) the set of fixed points of \(S\). The variational inequality problem \(VI(C,A)\) is to find \(u \in C\) such that \[ (Au,v-u)\geq 0 \quad \forall v \in C. \] Following conditions are used to express the main results: {\parindent=8mm \begin{itemize}\item[(A1)] \(\{\sigma_{n} \}\subset (0,1/k)\), \(\{\varepsilon_{n}\}\subset [0,\infty)\) and \(\sum_{n=1}^{\infty} \varepsilon_{n} <\infty\);\item[(A2)] \(\{\lambda_{n}\}\subset [a,b]\) for some \(a,b \in (0,1/k)\); \item[(A3)] \(\{\alpha_{n}\}\subset [0,c]\) for some \(c \in [0,1)\). \end{itemize}} The authors study the following algorithms: I. Let \(F(S)\cap VI(C,A)\) be bounded. Starting with an arbitrary initial guess \(x_{1}\in C\) generate three sequences \(\{x_{n}\},\{y_{n}\},\{z_{n}\}\) via the following implicit scheme: \[ \begin{cases} y_{n}=P_{C}(x_{n}-\lambda_{n}\mu Ax_{n}-\lambda_{n}(1-\mu)Ay_{n}), \\ z_{n}=\alpha_{n} x_{n}+ (1-\alpha_{n})S^{n}P_{C}(x_{n}-\lambda_{n}Ay_{n}),\\ C_{n}=\{z \in C:\| z_{n}-z\|^{2}\leq\|x_{n}-z\|^{2}+\theta_{n}\},\end{cases} \] find \(x_{n+1}\in C_{n}\) such that \((x_{n}-x_{n-1}+e_{n}-\sigma_{n}Ax_{n+1}, x_{n+1}-x)\geq -\varepsilon_{n}\) \(\forall x\in C_{n}\) and \(n\in {\mathbb N}\). It is shown that if conditions (A1)-(A3) are fulfilled the sequences \({x_{n}},{y_{n}},{z_{n}}\) generated by this implicit scheme converge to the same element of \(F(S)\bigcap VI(C,A)\). In algorithm II \(C\) is supposed too be bounded the sequences being generated in the same way.