Contraining Strategies for Linear Iterative Processes

Abstract
Solutions of linear systems of equations are often known in advance to belong to some closed convex set B⊆Rn. This information leads to certain constraining strategies in iterative solution algorithms, such as the (nonlinear) projection of iterates onto B. Convergence of resulting iterative processes and their limits are studied.

This publication has 0 references indexed in Scilit: