Markov Chain Monte Carlo Methods for Hierarchical Bayesian Expert Systems

  • Jeremy C. York
  • David Madigan
Part of the Lecture Notes in Statistics book series (LNS, volume 89)


In a hierarchical Bayesian expert system, the probabilities relating the variables are not known precisely; rather, imprecise knowledge of these probabilities is described by placing prior distributions on them. After obtaining data, one would like to update those distributions to reflect the new information gained; however, this can prove difficult computationally if the observed data are incomplete. This paper describes a way around these difficulties—use of Markov chain Monte Carlo methods.


Markov Chain Monte Carlo Gibbs Sampler Markov Chain Monte Carlo Method Markov Chain Monte Carlo Approach Markov Chain Monte Carlo Scheme 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [Cooper 84]
    Cooper, G. F. (1984) NESTOR: A computer-based medical diagnostic aid that integrates causal and probabilistic knowledge, PhD Thesis, Dept of Computer Science, Stanford University.Google Scholar
  2. [Cooper 92]
    Cooper, G. F. (1992) Personal communication.Google Scholar
  3. [DawiLaur 89]
    Dawid, A.P. and Lauritzen, S.L. (1989)“Markov Distributions, Hyper-Markov Laws and Meta-Markov Models on Decomposable Graphs, with Applications to Bayesian Learning in Expert Systems,” Technical Report R-89-31. Institute for Electronic Systems, Aalborg University.Google Scholar
  4. [DawiLaur 93]
    Dawid, A.P. and Lauritzen, S.L. (1993) “Hyper Markov Laws in the Statistical Analysis of Decomposable Graphical Models,” Ann. Stat., in pressGoogle Scholar
  5. [Devroye 86]
    Devroye, Luc (1986) Non-uniform Random Variate Generation. Springer-Verlag, New York.MATHGoogle Scholar
  6. [GelfSmit 91]
    Gelfand, A. and Smith, A.F.M. (1991) “Gibbs sampling for marginal posterior expectations,” Communs. Statist. Theory Meth. 20, 1747–1766.MathSciNetCrossRefGoogle Scholar
  7. [Geyer 92]
    Geyer, Charles J. (1992) “Practical Markov Chain Monte Carlo,” Stat. Sci. 7, 473–483.CrossRefGoogle Scholar
  8. [GeyeTier 92]
    Geyer, Charles J. and Tierney, Luke (1992) “On the Convergence of Monte Carlo Approximations to the Posterior Density,” Technical Report 579. School of Statistics, University of Minnesota.Google Scholar
  9. [Hastings 70]
    Hastings, W.K. (1970) “Monte Carlo Sampling Methods Using Markov Chains and Their Applications,” Biometrika 57, 97 - 109.MATHCrossRefGoogle Scholar
  10. [LauDawLarLei 90]
    Lauritzen, S.L., Dawid, A.P., Larsen, B. N. and Leimer, H.G. (1990) “Independence Properties of Directed Markov Random Fields,” Networks 20, 491 - 505.MathSciNetMATHCrossRefGoogle Scholar
  11. [LaurSpie 88]
    Lauritzen, S.L. and Spiegelhalter, D. (1988) “Local Computations with Probabilities on Graphical Structures and their Application to Expert Systems (with Discussion),” J. Roy. Statist. Soc. ser. B 50, 157 - 224.MathSciNetMATHGoogle Scholar
  12. [MadiYork 93]
    Madigan, D. and York, Jeremy C. (1993) “Bayesian Graphical Models,” Technical Report 259. Department of Statistics, University of Washington.Google Scholar
  13. [Pearl 88]
    Pearl, J. (1988) Probabilistic Reasoning in Intelligent Systems. Morgan Kaufmann, San Mateo.Google Scholar
  14. [Ripley 87]
    Ripley, B. (1987) Stochastic Simulation. John Wiley and Sons, New York.MATHCrossRefGoogle Scholar
  15. [SmitRobe 93]
    Smith, A.F.M. and Roberts, G. O. (1993) “Bayesian Computation via the Gibbs Sampler and Related Markov Chain Monte Carlo Methods,” J. R. Statist. Soc. B 55, 3–23.MathSciNetMATHGoogle Scholar
  16. [SpieCowe 92]
    Spiegelhalter, D.J. and Cowell, R.G. (1992) “Learning in Probabilistic Expert Systems,” in: Bayesian Statistics 4, Bernardo, J.M., Berger, J.O., Dawid, A.P. and Smith, A.F.M., eds, Oxford University Press, Oxford.Google Scholar
  17. [SpieLaur90]
    Spiegelhalter, D.J. and Lauritzen, S.L. (1990) “Sequential Updating of Conditional Probabilities on Directed Graphical Structures,” Networks 20, 579–605.MathSciNetMATHCrossRefGoogle Scholar
  18. [Tierney 91a]
    Tierney, Luke ( 1991 a) “Markov Chains for Exploring Posterior Distributions.” Technical Report 560, School of Statistics, University of Minnesota.Google Scholar
  19. [Tierney 91b]
    Tierney, Luke (1991b) “Exploring Posterior Distributions Using Markov Chains,” in: Computing Science and Statistics, Proceedings of the 23rd Symposium on the Interface, 563–570.Google Scholar
  20. [WeiTann 90]
    Wei, G.C.G. and Tanner, M. A. (1990) “Calculating the Content and the Boundary of the Highest Posterior Density Region via Data Augmentation”, Biometrika 77, 649–652.MathSciNetCrossRefGoogle Scholar
  21. [York 92a]
    York, Jeremy C. (1992a) “Use of the Gibbs sampler in Expert Systems,” Artificial Intelligence 56, 115 - 130.MathSciNetMATHCrossRefGoogle Scholar
  22. [York 92b]
    York, Jeremy C. (1992b) Bayesian Methods for the Analysis of Misclassified or Incomplete Multivariate Discrete Data, PhD Thesis, Dept of Statistics, University of Washington.Google Scholar
  23. [YorkMadi 92]
    “Bayesian methods for estimating the size of a closed population.” Technical report 234, Department of Statistics, University of Washington.Google Scholar
  24. [YorMadHeuLie 93]
    York, Jeremy C., Madigan, David, Heuch, Ivar and Lie, Rolv Terje (1993) “Birth Defects Registered by Double Sampling: A Bayesian Approach Incorporating Covariates and Model Uncertainty,” submitted for publication.Google Scholar

Copyright information

© Springer-Verlag New York 1994

Authors and Affiliations

  • Jeremy C. York
    • 1
  • David Madigan
    • 2
  1. 1.Dept of StatisticsCarnegie Mellon UniversityUSA
  2. 2.Dept of StatisticsUniversity of WashingtonUSA

Personalised recommendations