Randomized and fault-tolerant method of subspace corrections

Abstract

In this paper, we consider the iterative method of subspace corrections with random ordering. We prove identities for the expected convergence rate and use these results to provide sharp estimates for the expected error reduction per iteration. We also study the fault-tolerant features of the randomized successive subspace correction method by rejecting corrections when faults occur and show that the resulting iterative method converges with probability one. In addition, we derive estimates on the expected convergence rate for the fault-tolerant, randomized, subspace correction method.