An iterative algorithm for large size least-squares constrained regularization problems (Q555485)

An iterative constrained least squares regularization (CLSRit) algorithm is defined and studied. It computes a sequence of solution approximations and parameters that converge (in absence of data noise) to the solution of a constrained optimization problem as a formulation for the regularization problem. In the case of noisy data the defined algorithm computes good regularized solutions at a low computational cost. An estimation procedure of the regularity of solution and suitable stopping criteria are introduced to improve the algorithm efficiency. The numerical experiments performed on test problem prove that CLSEit algorithm is an efficient method that can be easily applied to large size regularization problems and show that the algorithm gives good results both in terms of precision and computational efficiency.