Newton iteration for partial differential equations and the approximation of the identity (Q5934360)

It is known that the critical condition guaranteeing quadratic convergence of approximate Newton methods is an approximation of the identity condition. This requires that the composition of the numerical inversion of the Frechet derivative with the derivative itself approximate the identity to an accuracy calibrated by the residual. It translates in numerical methods to a polynomial function of the grid size, chosen adaptively according to the current residual. The reviewed article deals with the effective implementation of approximate Newton methods for a differential map, defined via a numerical discretization. The authors study the approximation of the identity condition with respect to dual norms and discuss the effect upon convergence. Connection to Nash iteration is also discussed.