Error estimates for iterative algorithms for minimizing regularized quadratic subproblems

We derive bounds for the objective errors and gradient residuals when finding approximations to the solution of common regularized quadratic optimization problems within evolving Krylov spaces. These provide upper bounds on the number of iterations required to achieve a given stated accuracy. We illustrate the quality of our bounds on given test examples.

Citation

Technical Report RAL-TR-2019-004, STFC-Rutherford Appleton Laboratory, Oxfordshire, England, March 2019

Article

Download

View PDF