Encyclopedia of Applied and Computational Mathematics

2015 Edition
| Editors: Björn Engquist

Bayesian Statistics: Computation

  • Martin A. Tanner
Reference work entry
DOI: https://doi.org/10.1007/978-3-540-70529-1_327

Mathematics Subject Classification

62F15; 65C40

Synonyms

Markov chain Monte Carlo (MCMC)

Short Definition

Markov chain Monte Carlo (MCMC) is a collection of computational methods for simulating from posterior distributions.

Description

Markov chain Monte Carlo (MCMC) methods are a collection of computational algorithms designed to sample from a target distribution by performing Monte Carlo simulation from a Markov chain whose equilibrium distribution is equal to the target distribution. The output of the algorithm is then used to estimate features of the required distribution, where the quality of the estimate is determined by the number of iterations of the algorithm. Surprisingly, it took several decades before the statistical community embraced Markov chain Monte Carlo (MCMC) as a general computational tool in Bayesian inference, where it may be quite difficult to compute the normalizing constant of the (possibly high-dimensional) posterior distribution required for routine...

This is a preview of subscription content, log in to check access.

References

  1. 1.
    Binder, K.: Monte Carlo Methods in Statistical Physics. Springer, New York (1978)Google Scholar
  2. 2.
    Dempster, A.P., Laird, N.M., Rubin, D.B.: Maximum likelihood from incomplete data via the EM algorithm (with discussion). J. R. Stat. Soc. B: Stat. Methodol. 39, 1–38 (1977)MathSciNetzbMATHGoogle Scholar
  3. 3.
    Gelfand, A.E., Smith, A.F.M.: Sampling-based approaches to calculating marginal densities. J. Am. Stat. Assoc. 85, 398–409 (1990)MathSciNetzbMATHCrossRefGoogle Scholar
  4. 4.
    Geman, S., Geman, D.: Stochastic relaxation, Gibbs distributions and the Bayesian restoration of images. IEEE Trans. Pattern Anal. Mach. Intell. 6, 721–741 (1984)zbMATHCrossRefGoogle Scholar
  5. 5.
    Hammersley, J.M., Handscomb, D.C.: Monte Carlo Methods, 2nd edn. Chapman and Hall, London (1964)zbMATHCrossRefGoogle Scholar
  6. 6.
    Hastings, W.K.: Monte Carlo sampling methods using Markov chains and their applications. Biometrika 57, 97–109 (1970)zbMATHCrossRefGoogle Scholar
  7. 7.
    Kloek, T., van Dijk, H.K.: Bayesian estimates of equation system parameters: an application of integration by Monte Carlo. Econometrica 46, 1–19 (1978)zbMATHCrossRefGoogle Scholar
  8. 8.
    Metropolis, N., Rosenbluth, A.W., Rosenbluth, M.N., Teller, A.H., Teller, E.: Equation of state calculations by fast computing machines. J. Chem. Phys. 21, 1087–1091 (1953)CrossRefGoogle Scholar
  9. 9.
    Morris, C.N.: Comment on “The calculation of posterior distributions by data augmentation” by M.A. Tanner and W.H. Wong. J. Am. Stat. Assoc. 82, 542–543 (1987)Google Scholar
  10. 10.
    O’Hagan, A.: Monte Carlo is fundamentally unsound. J. R. Stat. Soc. D: Stat. 36, 247–249 (1987)Google Scholar
  11. 11.
    Smith, A.F.M., Skene, A.M., Shaw, J.E.H., Naylor, J.C., Dransfield, M.: The implementation of the Bayesian paradigm. Commun. Stat. Theory Methods 14(5), 1079–1102 (1985)MathSciNetzbMATHCrossRefGoogle Scholar
  12. 12.
    Swendsen, R.H., Wang, J.S.: Nonuniversal critical dynamics in Monte Carlo simulations. Phys. Rev. Lett. 58, 86–88 (1987)CrossRefGoogle Scholar
  13. 13.
    Tanner, M.A.: Metropolis algorithms. In: The Encyclopedia of Applied and Computational Mathematics (2014)Google Scholar
  14. 14.
    Tanner, M.A., Wong, W.H.: The calculation of posterior distributions by data augmentation (with discussion). J. Am. Stat. Assoc. 82, 528–550 (1987)MathSciNetzbMATHCrossRefGoogle Scholar
  15. 15.
    Tanner, M.A., Wong, W.H.: From EM to data augmentation: the emergence of MCMC Bayesian computation in the 1980s. Stat. Sci. 25, 506–516 (2010)MathSciNetzbMATHCrossRefGoogle Scholar
  16. 16.
    Tierney, L., Kadane, J.B.: Accurate approximations for posterior moments and marginal densities. J. Am. Stat. Assoc. 81, 82–86 (1986)MathSciNetzbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2015

Authors and Affiliations

  • Martin A. Tanner
    • 1
  1. 1.Department of StatisticsNorthwestern UniversityEvanstonUSA