Electronic Journal of Statistics

Construction of weakly CUD sequences for MCMC sampling

Seth D. Tribble and Art B. Owen

Full-text: Open access

Abstract

In Markov chain Monte Carlo (MCMC) sampling considerable thought goes into constructing random transitions. But those transitions are almost always driven by a simulated IID sequence. Recently it has been shown that replacing an IID sequence by a weakly completely uniformly distributed (WCUD) sequence leads to consistent estimation in finite state spaces. Unfortunately, few WCUD sequences are known. This paper gives general methods for proving that a sequence is WCUD, shows that some specific sequences are WCUD, and shows that certain operations on WCUD sequences yield new WCUD sequences. A numerical example on a 42 dimensional continuous Gibbs sampler found that some WCUD inputs sequences produced variance reductions ranging from tens to hundreds for posterior means of the parameters, compared to IID inputs.

Article information

Source
Electron. J. Statist., Volume 2 (2008), 634-660.

Dates
First available in Project Euclid: 30 July 2008

Permanent link to this document
https://projecteuclid.org/euclid.ejs/1217450798

Digital Object Identifier
doi:10.1214/07-EJS162

Mathematical Reviews number (MathSciNet)
MR2426105

Zentralblatt MATH identifier
1320.62055

Subjects
Primary: 62F15: Bayesian inference
Secondary: 11K45: Pseudo-random numbers; Monte Carlo methods 11K41: Continuous, $p$-adic and abstract analogues

Keywords
completely uniformly distributed Gibbs sampler equidistribution probit quasi-Monte Carlo

Citation

Tribble, Seth D.; Owen, Art B. Construction of weakly CUD sequences for MCMC sampling. Electron. J. Statist. 2 (2008), 634--660. doi:10.1214/07-EJS162. https://projecteuclid.org/euclid.ejs/1217450798


Export citation

References

  • Albert, J. and Chib, S. (1993). Bayesian analysis of binary and polychotomous response data., Journal of the American Statistical Association, 88:669–679.
  • Blair, J. M., Edwards, C. A., and Johnson, J. H. (1976). Rational Chebyshev approximation for the inverse of the error function., Mathematics of Computation, 30(136):827–830.
  • Caflisch, R. E., Morokoff, W., and Owen, A. B. (1997). Valuation of mortgage backed securities using Brownian bridges to reduce effective dimension., Journal of Computational Finance, 1:27–46.
  • Chaudary, S. (2004)., Acceleration of Monte Carlo methods using low discrepancy sequences. PhD thesis, UCLA.
  • Chentsov, N. (1967). Pseudorandom numbers for modelling Markov chains., Computational Mathematics and Mathematical Physics, 7:218–2332.
  • Craiu, R. V. and Lemieux, C. (2007). Acceleration of multiple-try Metropolis algorithm using antithetic and stratified sampling., Statistics and computing, 17(2).
  • Cranley, R. and Patterson, T. (1976). Randomization of number theoretic methods for multiple integration., SIAM Journal of Numerical Analysis, 13:904–914.
  • Devroye, L. (1986)., Non-uniform Random Variate Generation. Springer.
  • Finney, D. J. (1947). The estimation from individual records of the relationship between dose and quantal response., Biometrika, 34:320–334.
  • Gilks, W., Richardson, S., and Spiegelhalter, D. (1996)., Markov Chain Monte Carlo in Practice. Chapman and Hall, Boca Raton.
  • Hörmann, W., Leydold, J., and Derflinger, G. (2004)., Automatic Nonuniform Random Variate Generation. Springer, Berlin.
  • Knuth, D. E. (1998)., The Art of Computer Programming, volume 2: Seminumerical algorithms. Addison-Wesley, Reading MA, Third edition.
  • Korobov, N. M. (1948). On functions with uniformly distributed fractional parts., Dokl. Akad. Nauk SSSR, 62:21–22. (In Russian).
  • Landau, D. P. and Binder, K. (2005)., A Guide to Monte Carlo Simulations in Statistical Physics. Cambridge University Press, New York, second edition.
  • L’Ecuyer, P. (1999). Tables of linear congruential generators of different sizes and good lattice structure., Mathematics of Computation, 68:249–260.
  • L’Ecuyer, P. (2008)., SSJ: A Java Library for Stochastic Simulation. Software user’s guide, Available at http://www.iro.umontreal.ca/~lecuyer.
  • L’Ecuyer, P., Lécot, C., and Tuffin, B. (2005). Randomized Quasi-Monte Carlo simulation of Markov chains with an ordered state space. In Niederreiter, H. and Talay, D., editors, Monte Carlo and Quasi-Monte Carlo Methods 2004. Springer. To appear.
  • Lemieux, C. and Sidorsky, P. (2005). Exact sampling with highly-uniform point sets. Technical, report.
  • Levin, M. (1999). Discrepancy estimates of completely uniformly distributed and pseudo-random number sequences., International Mathematics Research Notices, pages 1231–1251.
  • Liao, L. G. (1998). Variance reduction in Gibbs sampler using quasi random numbers., Journal of Computational and Graphical Statistics, 7:253–266.
  • Liu, J. S. (2001)., Monte Carlo strategies in scientific computing. Springer, New York.
  • Liu, J. S., Liang, F., and Wong, W. H. (2000). The use of multiple-try method and local optimization in Metropolis sampling., Journal of the American Statistical Association, 95:121–134.
  • Newman, M. E. J. and Barkema, G. T. (1999)., Monte Carlo Methods in Statistical Physics. Oxford University Press, New York.
  • Niederreiter, H. (1977). Pseudo-random numbers and optimal coefficients., Advances in Mathematics, 26:99–181.
  • Niederreiter, H. (1992)., Random Number Generation and Quasi-Monte Carlo Methods. S.I.A.M., Philadelphia, PA.
  • Owen, A. B. (2005). Multidimensional variation for quasi-Monte Carlo. In Fan, J. and Li, G., editors, International Conference on Statistics in honour of Professor Kai-Tai Fang’s 65th birthday.
  • Owen, A. B. and Tribble, S. D. (2005). A quasi-Monte Carlo Metropolis algorithm., Proceedings of the National Academy of Science, 102(25):8844–8849.
  • Propp, J. G. and Wilson, D. B. (1996). Exact sampling with coupled Markov chains and applications to statistical mechanics., Random Structures and Algorithms, 9:223–252.
  • Robert, C. and Casella, G. (2004)., Monte Carlo Statistical Methods. Springer, New York, 2nd edition.
  • Sloan, I. H. and Joe, S. (1994)., Lattice Methods for Multiple Integration. Oxford Science Publications, Oxford.
  • Tribble, S. D. (2007)., Markov chain Monte Carlo algorithms using completely uniformly distributed driving sequences. PhD thesis, Stanford University.