Additive preconditioning for matrix computations (Q846341)

Randomized additive preconditioners which are often used to solve linear systems of equations for a very large class of input matrices are considered. The generation of such preconditioners and their impact on the rank and the condition number of a matrix are investigated. Two alternative versions of randomized preprocessing are proposed. Weak randomization is employed to construct structured preconditioners, defined by a small number of random parameters. The numerical tests are performed to confirm the basic results.