On multi-step greedy randomized coordinate descent method for solving large linear least-squares problems

摘要

To solve large-scale linear least-squares problems, we propose a multi-step greedy randomized coordinate descent method based on the greedy randomized coordinate descent method. We also prove that the new method converges to the unique solution of the linear least-squares problem when the coefficient matrix is of full rank, and the number of rows is not less than the number of columns. Finally, some numerical experiments demonstrate the effectiveness of the multi-step greedy randomized coordinate descent method in solving linear least-squares problems.

关键词