Abstract
Polynomial acceleration methods from computational optimization can be applied to accelerating MCMC. For example, a geometrically convergent MCMC may be accelerated to be a perfect sampler in special circumstances. An equivalence between Gibbs sampling of Gaussian distributions and classical iterative methods can be established using matrix splittings, allowing direct application of Chebyshev acceleration. The conjugate gradient method can also be adapted to give an accelerated sampler for Gaussian distributions, that is perfect in exact arithmetic.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Adler, S.L.: Over-relaxation method for the Monte Carlo evaluation of the partition function for multiquadratic actions. Phys. Rev. D 23, 2901–2904 (1981)
Amit, Y., Grenander, U.: Comparing sweep strategies for stochastic relaxation. J. Multivariate Anal. 37, 197–222 (1991)
Axelsson, O.: Iterative Solution Methods. Cambridge University Press, Cambridge (1996)
Barone, P., Frigessi, A.: Improving stochastic relaxation for Gaussian random fields. Probab. Engrg. Inform. Sci. 4, 369–384 (1990)
Ceriotti, M., Bussi, G., Parrinello, M.: Conjugate gradient heat bath for ill-conditioned actions. Phys. Rev. E 76, 026707-1–8 (2007)
Diaconis, P., Freedman, D.: Iterated random functions. SIAM Rev. 41, 45–76 (1999)
Fox, C.: A conjugate direction sampler for normal distributions, with a few computed examples. Technical Reports from the Electronics Group, University of Otago (2008)
Fox, C., Parker, A.: Convergence in variance of Chebyshev accelerated Gibbs samplers. SIAM J. Sci. Comput. (2013, to appear)
Fox, C., Parker, A.: Gibbs sampling of normal distributions using matrix splittings and polynomial acceleration. (2013, in preparation)
Galli, A., Gao, H.: Rate of convergence of the Gibbs sampler in the Gaussian case. Mathematical Geology 33, 653–677 (2001)
Geman, S., Geman, D.: Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images. IEEE Trans. Pattern Anal. Mach. Intell. 6, 721–741 (1984)
Glauber, R.: Time dependent statistics of the Ising model. J. Math. Phys. 4, 294–307 (1963)
Golub, G.H., Varga, R.S.: Chebyshev semi-iterative methods, successive over-relaxation iterative methods, and second-order Richardson iterative methods, parts I and II. Numer. Math. 3, 147–156, 157–168 (1961)
Goodman, J., Sokal, A.D.: Multigrid Monte Carlo method. Conceptual foundations. Phys. Rev. D 40, 2035–2071 (1989)
Higdon, D.: A primer on space-time modelling from a Bayesian perspective. In: Finkenstadt, B., Held, L., Isham, V. (eds.) Statistics of Spatio-Temporal Systems, pp. 217–279. Chapman & Hall/CRC, New York (2006)
Łatuszyński, K., Roberts, G.O.: CLTs and asymptotic variance of time-sampled Markov chains. Methodol. Comput. Appl. Probab. 15, 237–247 (2013)
Parker, A., Fox, C.: Sampling Gaussian distributions in Krylov spaces with conjugate gradients. SIAM J. Sci. Comput. 34, B312–B334 (2012)
Roberts, G.O., Sahu, S.K.: Updating schemes, correlation structure, blocking and parameterization for the Gibbs sampler. J. R. Stat. Soc. Ser. B. 59, 291–317 (1997)
Rue, H.: Fast sampling of Gaussian Markov random fields. J. R. Stat. Soc. Ser. B. 63, 325–338 (2001)
Schneider, M.K., Willsky, A.S.: A Krylov subspace method for covariance approximation and simulation of random processes and fields. Multidimens. Syst. Signal Process. 14, 295–318 (2003)
Turchin, V.F.: On the computation of multidimensional integrals by the Monte Carlo method. Theory Probab. Appl. 16, 720–724 (1971)
Young, D.M.: Iterative Solution of Large Linear Systems. Academic, New York (1971)
Acknowledgements
Polynomial acceleration of Gibbs sampling is the brainchild of Al Parker, to whom I am indebted. This research was supported by Marsden contract UOO1015.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Fox, C. (2013). Polynomial Accelerated MCMC and Other Sampling Algorithms Inspired by Computational Optimization. In: Dick, J., Kuo, F., Peters, G., Sloan, I. (eds) Monte Carlo and Quasi-Monte Carlo Methods 2012. Springer Proceedings in Mathematics & Statistics, vol 65. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-41095-6_15
Download citation
DOI: https://doi.org/10.1007/978-3-642-41095-6_15
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-41094-9
Online ISBN: 978-3-642-41095-6
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)