scientific article; zbMATH DE number 2068406

From MaRDI portal
Publication:4464979





The Newton method is considered for approximating locally unique solutions of the nonlinear operator equation \(F(x)=0\) and its discretization \(P_h(y)=0\). Here \(F\) is a Fréchet-differentiable operator from the Banach space \(X\) into \(Y\), and \(P_h\) is a discretization of \(F\). The mesh independence principal states that the Newton method for the continuous and discrete problems differs at most by one in the number of steps required to converge to within a given tolerance. Using a combination of Lipschitz and center-Lipschitz conditions: \[ \| F'(x_0)^{-1}[F'(x)-F'(x_0)]\| \leq b\| x-x_0\| \quad\text{and}\quad \| F'(x_0)^{-1}[F(x)-F(y)]\| \leq a\| x-y\| , \] the author improves the error estimate, radius of convergence and number of iterations in the previous study of the mesh independence principal.











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 Q4464979)