Statistical Science

Component-Wise Markov Chain Monte Carlo: Uniform and Geometric Ergodicity under Mixing and Composition

Alicia A. Johnson, Galin L. Jones, and Ronald C. Neath

Full-text: Open access

Abstract

It is common practice in Markov chain Monte Carlo to update the simulation one variable (or sub-block of variables) at a time, rather than conduct a single full-dimensional update. When it is possible to draw from each full-conditional distribution associated with the target this is just a Gibbs sampler. Often at least one of the Gibbs updates is replaced with a Metropolis–Hastings step, yielding a Metropolis–Hastings-within-Gibbs algorithm. Strategies for combining component-wise updates include composition, random sequence and random scans. While these strategies can ease MCMC implementation and produce superior empirical performance compared to full-dimensional updates, the theoretical convergence properties of the associated Markov chains have received limited attention. We present conditions under which some component-wise Markov chains converge to the stationary distribution at a geometric rate. We pay particular attention to the connections between the convergence rates of the various component-wise strategies. This is important since it ensures the existence of tools that an MCMC practitioner can use to be as confident in the simulation results as if they were based on independent and identically distributed samples. We illustrate our results in two examples including a hierarchical linear mixed model and one involving maximum likelihood estimation for mixed models.

Article information

Source
Statist. Sci., Volume 28, Number 3 (2013), 360-375.

Dates
First available in Project Euclid: 28 August 2013

Permanent link to this document
https://projecteuclid.org/euclid.ss/1377696941

Digital Object Identifier
doi:10.1214/13-STS423

Mathematical Reviews number (MathSciNet)
MR3135537

Zentralblatt MATH identifier
1331.60151

Keywords
Geometric ergodicity uniform ergodicity Markov chain Monte Carlo Gibbs sampler Metropolis-within-Gibbs random scan convergence rate

Citation

Johnson, Alicia A.; Jones, Galin L.; Neath, Ronald C. Component-Wise Markov Chain Monte Carlo: Uniform and Geometric Ergodicity under Mixing and Composition. Statist. Sci. 28 (2013), no. 3, 360--375. doi:10.1214/13-STS423. https://projecteuclid.org/euclid.ss/1377696941


Export citation

References

  • Atchadé, Y. F. (2011). Kernel estimators of asymptotic variance for adaptive Markov chain Monte Carlo. Ann. Statist. 39 990–1011.
  • Bédard, M. and Rosenthal, J. S. (2008). Optimal scaling of Metropolis algorithms: Heading toward general target distributions. Canad. J. Statist. 36 483–503.
  • Booth, J. G. and Hobert, J. P. (1999). Maximizing generalized linear mixed model likelihoods with an automated Monte Carlo EM algorithm. J. R. Stat. Soc. Ser. B Stat. Methodol. 61 265–285.
  • Caffo, B. S., Jank, W. and Jones, G. L. (2005). Ascent-based Monte Carlo expectation-maximization. J. R. Stat. Soc. Ser. B Stat. Methodol. 67 235–251.
  • Christensen, O. F., Møller, J. and Waagepetersen, R. P. (2001). Geometric ergodicity of Metropolis–Hastings algorithms for conditional simulation in generalized linear mixed models. Methodol. Comput. Appl. Probab. 3 309–327.
  • Coull, B. A., Hobert, J. P., Ryan, L. M. and Holmes, L. B. (2001). Crossed random effect models for multiple outcomes in a study of teratogenesis. J. Amer. Statist. Assoc. 96 1194–1204.
  • Diaconis, P., Khare, K. and Saloff-Coste, L. (2008). Gibbs sampling, exponential families and orthogonal polynomials. Statist. Sci. 23 151–178.
  • Doss, H. and Hobert, J. P. (2010). Estimation of Bayes factors in a class of hierarchical random effects models using a geometrically ergodic MCMC algorithm. J. Comput. Graph. Statist. 19 295–312.
  • Flegal, J. M., Haran, M. and Jones, G. L. (2008). Markov chain Monte Carlo: Can we trust the third significant figure? Statist. Sci. 23 250–260.
  • Flegal, J. M. and Jones, G. L. (2010). Batch means and spectral variance estimators in Markov chain Monte Carlo. Ann. Statist. 38 1034–1070.
  • Flegal, J. M. and Jones, G. L. (2011). Implementing Markov chain Monte Carlo: Estimating with confidence. In Handbook of Markov Chain Monte Carlo (S. P. Brooks, A. Gelman, G. L. Jones and X. L. Meng, eds.). CRC Press, Boca Raton, FL.
  • Fort, G., Moulines, E., Roberts, G. O. and Rosenthal, J. S. (2003). On the geometric ergodicity of hybrid samplers. J. Appl. Probab. 40 123–146.
  • Geyer, C. (1999). Likelihood inference for spatial point processes. In Stochastic Geometry (Toulouse, 1996) (O. E. Barndorff-Nielsen, W. S. Kendall and M. N. M. van Lieshout, eds.). Monographs on Statistics and Applied Probability 80 79–140. Chapman & Hall/CRC, Boca Raton, FL.
  • Hobert, J. P. (2000). Hierarchical models: A current computational perspective. J. Amer. Statist. Assoc. 95 1312–1316.
  • Hobert, J. P. (2011). The data augmentation algorithm: Theory and methodology. In Handbook of Markov Chain Monte Carlo (S. P. Brooks, A. Gelman, G. L. Jones and X. L. Meng, eds.). CRC Press, Boca Raton, FL.
  • Hobert, J. P. and Geyer, C. J. (1998). Geometric ergodicity of Gibbs and block Gibbs samplers for a hierarchical random effects model. J. Multivariate Anal. 67 414–430.
  • Hobert, J. P., Jones, G. L., Presnell, B. and Rosenthal, J. S. (2002). On the applicability of regenerative simulation in Markov chain Monte Carlo. Biometrika 89 731–743.
  • Jarner, S. F. and Hansen, E. (2000). Geometric ergodicity of Metropolis algorithms. Stochastic Process. Appl. 85 341–361.
  • Johnson, L. T. and Geyer, C. J. (2012). Variable transformation to obtain geometric ergodicity in the random-walk Metropolis algorithm. Ann. Statist. 40 3050–3076.
  • Johnson, A. A. and Jones, G. L. (2010). Gibbs sampling for a Bayesian hierarchical general linear model. Electron. J. Stat. 4 313–333.
  • Johnson, A. A., Jones, G. L. and Neath, R. C. (2013). Supplement to “Component-Wise Markov Chain Monte Carlo: Uniform and Geometric Ergodicity Under Mixing and Composition.” DOI:10.1214/13-STS423SUPP.
  • Jones, G. L. and Hobert, J. P. (2001). Honest exploration of intractable probability distributions via Markov chain Monte Carlo. Statist. Sci. 16 312–334.
  • Jones, G. L. and Hobert, J. P. (2004). Sufficient burn-in for Gibbs samplers for a hierarchical random effects model. Ann. Statist. 32 784–817.
  • Jones, G. L., Roberts, G. O. and Rosenthal, J. S. (2013). Convergence of conditional Metropolis–Hastings samplers. J. Appl. Probab. To appear.
  • Jones, G. L., Haran, M., Caffo, B. S. and Neath, R. (2006). Fixed-width output analysis for Markov chain Monte Carlo. J. Amer. Statist. Assoc. 101 1537–1547.
  • Łatuszynski, K., Roberts, G. O. and Rosenthal, J. S. (2013). Adaptive Gibbs samplers and related MCMC methods. Ann. Appl. Probab. 23 66–98.
  • Lee, K. J., Jones, G. L., Caffo, B. S. and Bassett, S. (2013). Spatial Bayesian variable selection models on functional magnetic resonance imaging time-series data. Preprint.
  • Marchev, D. and Hobert, J. P. (2004). Geometric ergodicity of van Dyk and Meng’s algorithm for the multivariate Student’s $t$ model. J. Amer. Statist. Assoc. 99 228–238.
  • McCulloch, C. E. (1997). Maximum likelihood algorithms for generalized linear mixed models. J. Amer. Statist. Assoc. 92 162–170.
  • Mengersen, K. L. and Tweedie, R. L. (1996). Rates of convergence of the Hastings and Metropolis algorithms. Ann. Statist. 24 101–121.
  • Meyn, S. P. and Tweedie, R. L. (1993). Markov Chains and Stochastic Stability. Springer, London.
  • Meyn, S. P. and Tweedie, R. L. (1994). Computable bounds for geometric convergence rates of Markov chains. Ann. Appl. Probab. 4 981–1011.
  • Neal, P. and Roberts, G. (2006). Optimal scaling for partially updating MCMC algorithms. Ann. Appl. Probab. 16 475–515.
  • Neath, R. C. (2013). On convergence properties of the Monte Carlo EM algorithm. In Advances in Modern Statistical Theory and Applications: A Festschrift in Honor of Morris L. Eaton (G. L. Jones and X. Shen, eds.). To appear.
  • Papaspiliopoulos, O. and Roberts, G. (2008). Stability of the Gibbs sampler for Bayesian hierarchical models. Ann. Statist. 36 95–117.
  • Robert, C. P. (1995). Convergence control methods for Markov chain Monte Carlo algorithms. Statist. Sci. 10 231–253.
  • Roberts, G. O. and Polson, N. G. (1994). On the geometric convergence of the Gibbs sampler. J. R. Stat. Soc. Ser. B Stat. Methodol. 56 377–384.
  • Roberts, G. O. and Rosenthal, J. S. (1997). Geometric ergodicity and hybrid Markov chains. Electron. Commun. Probab. 2 13–25 (electronic).
  • Roberts, G. O. and Rosenthal, J. S. (1998). Two convergence properties of hybrid samplers. Ann. Appl. Probab. 8 397–407.
  • Roberts, G. O. and Rosenthal, J. S. (1999). Convergence of slice sampler Markov chains. J. R. Stat. Soc. Ser. B Stat. Methodol. 61 643–660.
  • Roberts, G. O. and Rosenthal, J. S. (2001). Markov chains and de-initializing processes. Scand. J. Stat. 28 489–504.
  • Roberts, G. O. and Rosenthal, J. S. (2004). General state space Markov chains and MCMC algorithms. Probab. Surv. 1 20–71.
  • Roberts, G. O. and Sahu, S. K. (1997). Updating schemes, correlation structure, blocking and parameterization for the Gibbs sampler. J. R. Stat. Soc. Ser. B Stat. Methodol. 59 291–317.
  • Roberts, G. O. and Tweedie, R. L. (1996). Geometric convergence and central limit theorems for multidimensional Hastings and Metropolis algorithms. Biometrika 83 95–110.
  • Román, J. C. (2012). Convergence analysis of block Gibbs samplers for Bayesian general linear mixed models. Ph.D. thesis, Dept. Statistics, Univ. Florida.
  • Román, J. C. and Hobert, J. P. (2012). Convergence analysis of the Gibbs sampler for Bayesian general linear mixed models with improper priors. Ann. Statist. 40 2823–2849.
  • Rosenthal, J. S. (1995). Minorization conditions and convergence rates for Markov chain Monte Carlo. J. Amer. Statist. Assoc. 90 558–566.
  • Rosenthal, J. S. (1996). Analysis of the Gibbs sampler for a model related to James–Stein estimators. Statist. Comput. 6 269–275.
  • Rosenthal, J. S. (2011). Optimal proposal distributions and adaptive MCMC. In Handbook of Markov Chain Monte Carlo (S. P. Brooks, A. Gelman, G. L. Jones and X. L. Meng, eds.). CRC Press, Boca Raton, FL.
  • Roy, V. and Hobert, J. P. (2007). Convergence rates and asymptotic standard errors for Markov chain Monte Carlo algorithms for Bayesian probit regression. J. R. Stat. Soc. Ser. B Stat. Methodol. 69 607–623.
  • Tan, A. and Hobert, J. P. (2009). Block Gibbs sampling for Bayesian random effects models with improper priors: Convergence and regeneration. J. Comput. Graph. Statist. 18 861–878.
  • Tan, A., Jones, G. L. and Hobert, J. P. (2013). On the geometric ergodicity of two-variable Gibbs samplers. In Advances in Modern Statistical Theory and Applications: A Festschrift in Honor of Morris L. Eaton (G. L. Jones and X. Shen, eds.). To appear.
  • Tanner, M. A. and Wong, W. H. (1987). The calculation of posterior distributions by data augmentation (with discussion). J. Amer. Statist. Assoc. 82 528–550.
  • Tierney, L. (1994). Markov chains for exploring posterior distributions (with discussion). Ann. Statist. 22 1701–1762.

Supplemental materials

  • Supplementary material: Supplementary material for “Component-Wise Markov Chain Monte Carlo: Uniform and Geometric Ergodicity Under Mixing and Composition”. This supplemental article includes all technical details and proofs for the above results.