1: \begin{abstract}

2: The computation of stationary distributions of Markov chains is an important task

3: in the simulation of stochastic models. The linear systems arising in such applications involve

4: non-symmetric M-matrices, making algebraic multigrid

5: methods a natural choice for solving these systems.

6: In this paper we investigate extensions and

7: improvements of the bootstrap algebraic multigrid framework for solving these systems.

8: This is achieved by reworking the bootstrap setup process to use singular vectors

9: instead of eigenvectors in constructing interpolation and restriction.

10: We formulate a result concerning the convergence speed of GMRES for singular systems and experimentally justify why rapid convergence of the proposed method can be expected. We demonstrate its fast convergence and the favorable scaling behaviour for various test problems.

11: 

12: % Our main focus is on the stochastic matrices arising in the computation of stationary distributions of irreducible Markov chains. We use the resulting multigrid method as a preconditioner for a Krylov subspace iteration, give a theoretical justification for its convergence and illustrate its nice scaling behavior by a wide variety of numerical examples.

13: 

14: \end{abstract}

15: