Consistency of Markov chain quasi-Monte Carlo on continuous state spaces

The random numbers driving Markov chain Monte Carlo (MCMC) simulation are usually modeled as independent U(0, 1) random variables. Tribble [Markov chain Monte Carlo algorithms using completely uniformly distributed driving sequences (2007) Stanford Univ.] reports substantial improvements when those random numbers are replaced by carefully balanced inputs from completely uniformly distributed sequences. The previous theoretical justification for using anything other than i.i.d. U(0, 1) points shows consistency for estimated means, but only applies for discrete stationary distributions. We extend those results to some MCMC algorithms for continuous stationary distributions. The main motivation is the search for quasi-Monte Carlo versions of MCMC. As a side benefit, the results also establish consistency for the usual method of using pseudo-random numbers in place of random ones.

[1]  H. Lebesgue,et al.  Intégrale, Longueur, Aire , 1902 .

[2]  Takuji Nishimura,et al.  Mersenne twister: a 623-dimensionally equidistributed uniform pseudo-random number generator , 1998, TOMC.

[3]  李幼升,et al.  Ph , 1989 .

[4]  D. J. Finney,et al.  The estimation from individual records of the relationship between dose and quantal response. , 1947, Biometrika.

[5]  Harald Niederreiter,et al.  Random number generation and Quasi-Monte Carlo methods , 1992, CBMS-NSF regional conference series in applied mathematics.

[6]  Art B Owen,et al.  A quasi-Monte Carlo Metropolis algorithm. , 2005, Proceedings of the National Academy of Sciences of the United States of America.

[7]  J. Liao,et al.  Variance Reduction in Gibbs Sampler Using Quasi Random Numbers , 1998 .

[8]  Josef Dick,et al.  On Quasi-Monte Carlo Rules Achieving Higher Order Convergence , 2009 .

[9]  Limit theorems for iterated random functions by regenerative methods , 2001 .

[10]  Pierre L'Ecuyer,et al.  A Randomized Quasi-Monte Carlo Simulation Method for Markov Chains , 2006, Oper. Res..

[11]  M. Rosenblatt.  Remarks on a Multivariate Transformation , 1952 .

[12]  P. L'Ecuyer,et al.  Quasi-Monte Carlo via linear shift-register sequences , 1999, WSC'99. 1999 Winter Simulation Conference Proceedings. 'Simulation - A Bridge to the Future' (Cat. No.99CH37038).

[13]  J. Marsden,et al.  Elementary classical analysis , 1974 .

[14]  Christiane Lemieux,et al.  Acceleration of the Multiple-Try Metropolis algorithm using antithetic and stratified sampling , 2007, Stat. Comput..

[15]  Andrew Gelman,et al.  Inference from Simulations and Monitoring Convergence , 2011 .

[16]  H. Weyl.  Über die Gleichverteilung von Zahlen mod. Eins , 1916 .

[17]  Radford M. Neal.  Slice Sampling , 2003, The Annals of Statistics.

[18]  L. Devroye.  Non-Uniform Random Variate Generation , 1986 .

[19]  Donald Ervin Knuth,et al.  The Art of Computer Programming , 1968 .

[20]  Donald E. Knuth,et al.  The art of computer programming. Vol.2: Seminumerical algorithms , 1981 .

[21]  David Bruce Wilson,et al.  Exact sampling with coupled Markov chains and applications to statistical mechanics , 1996, Random Struct. Algorithms.

[22]  Christian P. Robert,et al.  Monte Carlo Statistical Methods , 2005, Springer Texts in Statistics.

[23]  Russel E. Caflisch,et al.  A quasi-Monte Carlo approach to particle simulation of the heat equation , 1993 .

[24]  R. Varga,et al.  Proof of Theorem 2 , 1983 .

[25]  Anand Srivastav,et al.  Finding optimal volume subintervals with k points and calculating the star discrepancy are NP-hard problems , 2009, J. Complex..

[26]  N. Chentsov.  Pseudorandom numbers for modelling Markov chains , 1967 .

[27]  Josef Dick.  A note on the existence of sequences with small star discrepancy , 2007, J. Complex..

[28]  Adrian F. M. Smith,et al.  Sampling-Based Approaches to Calculating Marginal Densities , 1990 .

[29]  S. Chib,et al.  Bayesian analysis of binary and polychotomous response data , 1993 .

[30]  Gareth O. Roberts,et al.  Convergence Properties of Perturbed Markov Chains , 1998, Journal of Applied Probability.

[31]  Persi Diaconis,et al.  Iterated Random Functions , 1999, SIAM Rev..

[32]  Michael C. Fu,et al.  Guest editorial , 2003, TOMC.

[33]  A. Owen.  Randomly Permuted (t,m,s)-Nets and (t, s)-Sequences , 1995 .

[34]  Felix Schlenk,et al.  Proof of Theorem 3 , 2005 .

[35]  Jun S. Liu,et al.  Monte Carlo strategies in scientific computing , 2001 .

[36]  Joshua N. Cooper,et al.  Deterministic Random Walks , 2006, ANALCO.

[37]  R. Bass,et al.  Review: P. Billingsley, Convergence of probability measures , 1971 .

[38]  Art B. Owen,et al.  Construction of weakly CUD sequences for MCMC sampling , 2008, 0807.4858.

[39]  D. Gaver,et al.  Robust empirical bayes analyses of event rates , 1987 .

[40]  C. Andrieu,et al.  On the ergodicity properties of some adaptive MCMC algorithms , 2006, math/0610317.

[41]  Christiane Lemieux,et al.  Exact sampling with highly uniform point sets , 2006, Math. Comput. Model..

[42]  J. Doob.  Stochastic processes , 1953 .

[43]  B. V. Limaye,et al.  A course in calculus and real analysis , 2006 .

[44]  H. Haario,et al.  An adaptive Metropolis algorithm , 2001 .

[45]  H. Niederreiter.  Multidimensional numerical integration using Pseudorandom numbers , 1986 .

[46]  Tobias Friedrich,et al.  Deterministic Random Walks on the Two-Dimensional Grid , 2009, Comb. Probab. Comput..

[47]  A. Owen.  Multidimensional variation for quasi-Monte Carlo , 2004 .

[48]  R. D. Gordon.  Values of Mills' Ratio of Area to Bounding Ordinate and of the Normal Probability Integral for Large Values of the Argument , 1941 .

[49]  R. Ash,et al.  Real analysis and probability , 1975 .

[50]  I. M. Sobol.  Pseudo-random numbers for constructing discrete Markov chains by the Monte Carlo method☆ , 1974 .