Advertisement

Statistics and Computing

, Volume 12, Issue 1, pp 17–26 | Cite as

Over-relaxation methods and coupled Markov chains for Monte Carlo simulation

  • Piero Barone
  • Giovanni Sebastiani
  • Julian Stander
Article

Abstract

This paper is concerned with improving the performance of certain Markov chain algorithms for Monte Carlo simulation. We propose a new algorithm for simulating from multivariate Gaussian densities. This algorithm combines ideas from coupled Markov chain methods and from an existing algorithm based only on over-relaxation. The rate of convergence of the proposed and existing algorithms can be measured in terms of the square of the spectral radius of certain matrices. We present examples in which the proposed algorithm converges faster than the existing algorithm and the Gibbs sampler. We also derive an expression for the asymptotic variance of any linear combination of the variables simulated by the proposed algorithm. We outline how the proposed algorithm can be extended to non-Gaussian densities.

coupled algorithms Gibbs sampler spectral radius 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Adler S.L. 1981. Over-relaxation method for the Monte Carlo evalua-tion of the partition function for multiquadratic actions. Physical Review D 23: 2901-2904.Google Scholar
  2. Barone P. and Frigessi A. 1990. Improving stochastic relaxation for Gaussian random fields. Probability in the Engineering and Informational Sciences 4: 369-389.Google Scholar
  3. Barone P., Sebastiani G., and Stander J. 1998. Metropolis coupled Markov chains and over-relaxation methods for Monte Carlo simulation.Istituto per le Applicazioni del Calcolo, CNR, Rome, Italy, Report 10/1998.Google Scholar
  4. Barone P., Sebastiani G., and Stander J. 2001. General over-relaxation Markov chain Monte Carlo algorithms for Gaussian densities.Statistics and Probability Letters, 52: 115-124.Google Scholar
  5. Brooks S.P. 1998. Markov chain Monte Carlo method and its applications.The Statistician 47: 69-100.Google Scholar
  6. Brooks S.P. and Roberts G.O. 1998. Convergence assessment techniques for Markov chain Monte Carlo. Statistics and Computing 8: 319-335.Google Scholar
  7. Cowles M.K. and Carlin B.P. 1996. Markov chain Monte Carlo convergence diagnostics: A comparative review. Journal of the American Statistical Association 91: 883-904.Google Scholar
  8. Foster M.A. and Hutchison J.M.S. (Eds.). 1987. Practical NMR Imaging.IRL Press Limited, Oxford.Google Scholar
  9. Geman S. and Geman D. 1984. Stochastic relaxation, Gibbs distribu-tions and the Bayesian restoration of images. IEEE Transactions on Pattern Analysis and Machine Intelligence PAMI-6: 721-741.Google Scholar
  10. Geyer C.J. 1991. Markov chain Monte Carlo maximum likelihood. In: Keramidas E.M. (Ed.), Computing Science and Statistics: Proceedings of the 23rd Symposium on the Interface. Interface Foundation, Fairfax Station, pp. 156-163.Google Scholar
  11. Gilks W.R., Richardson S., and Spiegelhalter, D.J. (Eds.). 1996. Markov Chain Monte Carlo in Practice. Chapman and Hall, London.Google Scholar
  12. Gilks W.R. and Roberts G.O. 1996. Strategies for improving MCMC.In: Gilks W.R., Richardson S., and Spiegelhalter D.J. (Eds.), Markov Chain Monte Carlo in Practice. Chapman and Hall, London, ch. 6, pp. 89-114.Google Scholar
  13. Golub G.H. and Van Loan C.F. 1989. Matrix Computations, 2nd edn. The Johns Hopkins University Press, Baltimore.Google Scholar
  14. Green P.J. and Han X.-L. 1992. Metropolis methods, Gaussian proposals and antithetic variables. In: Barone P., Frigessi A., and Piccioni M.(Eds.), Stochastic Models, Statistical Methods, and Algorithms in Image Analysis. Springer-Verlag, Berlin, pp. 142-164.Google Scholar
  15. Hastings W.K. 1970. Monte Carlo sampling methods using Markov chains and their applications. Biometrika 57: 97-109.Google Scholar
  16. Mengersen K.L., Robert C.P., and Guihenneuc-Jouyaux Ch. 1999.MCMC convergence diagnostics: A review ww. In: Bernardo J.M., Berger J.O., Dawid A.P., and Smith A.F.M. (Eds.), Bayesian Statistics 6. Oxford University Press, Oxford, pp. 415-440.Google Scholar
  17. Metropolis N. Rosenbluth A.W., Rosenbluth M.N., Teller A.H., and Teller E. 1953. Equations of state calculations by fast computing machines. Journal of Chemical Physics 21: 1087-1091.Google Scholar
  18. Neal R.M. 1995. Suppressing random walks in Markov chain Monte Carlo using ordered overrelaxation. Department of Statistics, University of Toronto, Technical Report 9508. Also available from http://www.statslab.cam.ac.uk/∼mcmc/Google Scholar
  19. Robert C. 1996. Convergence assessments for Markov chain Monte Carlo methods. Statistical Science 10: 231-253.Google Scholar
  20. Roberts G.O. and Sahu S.K. 1997. Updating schemes, correlation structure, blocking and parameterization for the Gibbs sampler. Journal of the Royal Statistical Society, Series B 59: 291-317.Google Scholar
  21. Rue H. 2001. Fast sampling of Gaussian Markov random fields. Journal of the Royal Statistical Society Series B 63: 325-338. Also available from http://www.statslab.cam.ac.uk/∼mcmc/Google Scholar
  22. Stewart G.W. 1973. Introduction to Matrix Computation. Academic Press, New York.Google Scholar
  23. Tierney L. 1994. Markov chains for exploring posterior distributions (with discussion). Annals of Statistics 22: 1701-1762.Google Scholar
  24. Tierney L. 1996. Introduction to general state-space Markov chain theory. In: Gilks W.R., Richardson S., and Spiegelhalter D.J.(Eds.), Markov Chain Monte Carlo in Practice. Chapman and Hall, London, ch. 4, pp. 59-88.Google Scholar
  25. Varga R.S. 2000. Matrix Iterative Analysis, 2nd ed. Springer-Verlag, Berlin.Google Scholar
  26. Wang Y. and Lee T. 1994. Statistical analysis of MR imaging and its applications in image modeling. In: Proc. IEEE Int. Conf. Image Processing Neural Networks, Vol. I, pp. 866-870.Google Scholar

Copyright information

© Kluwer Academic Publishers 2002

Authors and Affiliations

  • Piero Barone
  • Giovanni Sebastiani
  • Julian Stander

There are no affiliations available

Personalised recommendations