scientific article; zbMATH DE number 13351

From MaRDI portal
Publication:3977936





The problem of improvements of iteration methods for the approximation of a solution \(x^*\) of the linear operator equation \(x=y+L(x)\) on a separable Hilbert space \(H\), where \(y\in H\) is a given element and \(L\) is a compact linear operator on \(H\) is considered. The two methods of iteration are compared. 1. Considering a suitable complete basis \(\{z_ j\}\), \(j=1,2,\dots,n\), of elements in \(H\) and choosing constants \(\{c_{j,k}\}\) such that \(\bar x_ i=\sum^ i_{j=1}c_{i,j}z_ j\) is an approximate solution of the given equation, for the approximation of this solution for example Galerkin methods or least squares methods may be used. 2. Considering iterated approximation of the form \(\bar x_ i=y+\sum^ i_{j=1}d_{i,j}L(z_ j)\). The main result of the paper is the answer ``yes for the Galerkin method for the question: Is it true that \(\bar x_ i\) is improved by an iteration if \(i\) is chosen sufficiently large? Some results are also extended to include the approximate solution of the quadratic equation \(w=y+B(w,w)\), where \(B\) is a bilinear operator in \(H\), \(y\in H\) is the fixed and \(w\in H\) is the unknown element.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3977936)