Hostname: page-component-cd9895bd7-8ctnn Total loading time: 0 Render date: 2025-01-01T17:10:25.406Z Has data issue: false hasContentIssue false
  • English
  • Français

Ergodicity of Markov chain Monte Carlo with reversible proposal

Part of: Markov processes Probabilistic methods, simulation and stochastic differential equations

Published online by Cambridge University Press:  22 June 2017

K. Kamatani
K. Kamatani*
Affiliation:
Osaka University
*
* Postal address: Graduate School of Engineering Science and Center for Mathematical Modeling and Data Science, Osaka University, 1-3 Machikaneyama-cho, Toyonaka, Osaka 560-8531, Japan. Email address: [email protected]
Get access Rights & Permissions [Opens in a new window]

Abstract

We describe the ergodic properties of some Metropolis–Hastings algorithms for heavy-tailed target distributions. The results of these algorithms are usually analyzed under a subgeometric ergodic framework, but we prove that the mixed preconditioned Crank–Nicolson (MpCN) algorithm has geometric ergodicity even for heavy-tailed target distributions. This useful property comes from the fact that, under a suitable transformation, the MpCN algorithm becomes a random-walk Metropolis algorithm.

Keywords

Markov chainergodicityMonte Carloregular variation

MSC classification

Primary: 65C05: Monte Carlo methods
Secondary: 65C40: Computational Markov chains 60J05: Discrete-time Markov processes on general state spaces
Type
Research Papers
Information
Journal of Applied Probability , Volume 54 , Issue 2 , June 2017 , pp. 638 - 654
Copyright
Copyright © Applied Probability Trust 2017 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

[1] Atchadé, Y. and Fort, G. (2010). Limit theorems for some adaptive MCMC algorithms with subgeometric kernels. Bernoulli 16, 116–154. Google Scholar
[2] Beskos, A., Roberts, G., Stuart, A. and Voss, J. (2008). MCMC methods for diffusion bridges. Stoch. Dyn. 8, 319–350. Google Scholar
[3] Bingham, N. H., Goldie, C. M. and Teugels, J. L. (1989). Regular Variation (Encyclopedia Math. Appl. 27). Cambridge University Press. Google Scholar
[4] Brooks, S., Gelman, A., Jones, G. L. and Meng, X.-L. (2011). Handbook of Markov Chain Monte Carlo. Chapman & Hall/CRC, Boca Raton, FL. CrossRefGoogle Scholar
[5] Chen, Y., Keyes, D., Law, K. J. H. and Ltaief, H. (2015). Accelerated dimension-independent adaptive Metropolis. Preprint. Available at https://arxiv.org/abs/1506.05741v1. Google Scholar
[6] Dellaportas, P. and Roberts, G. O. (2003). An introduction to MCMC. In Spatial Statistics and Computational Methods (Aalborg, 2001; Lecture Notes Statist. 173), Springer, New York, pp. 1–41. Google Scholar
[7] Douc, R., Fort, G., Moulines, E. and Soulier, P. (2004). Practical drift conditions for subgeometric rates of convergence. Ann. Appl. Prob. 14, 1353–1377. CrossRefGoogle Scholar
[8] Dutta, S. (2012). Multiplicative random walk Metropolis-Hastings on the real line. Sankhyā B 74, 315–342. Google Scholar
[9] Fort, G. and Moulines, E. (2000). V-subgeometric ergodicity for a Hastings-Metropolis algorithm. Statist. Prob. Lett. 49, 401–410. CrossRefGoogle Scholar
[10] Fort, G. and Moulines, E. (2003). Polynomial ergodicity of Markov transition kernels. Stoch. Process. Appl. 103, 57–99. CrossRefGoogle Scholar
[11] Hairer, M., Stuart, A. M. and Vollmer, S. J. (2014). Spectral gaps for a Metropolis–Hastings algorithm in infinite dimensions. Ann. Appl. Prob. 24, 2455–2490. Google Scholar
[12] Hu, Z., Yao, Z. and Li, J. (2017). On an adaptive preconditioned Crank-Nicolson MCMC algorithm for infinite dimensional Bayesian inference. J. Comput. Phys. 332, 492–503. Google Scholar
[13] Jarner, S. F. and Hansen, E. (2000). Geometric ergodicity of Metropolis algorithms. Stoch. Process. Appl. 85, 341–361. Google Scholar
[14] Jarner, S. F. and Roberts, G. O. (2002). Polynomial convergence rates of Markov chains. Ann. Appl. Prob. 12, 224–247. Google Scholar
[15] Jarner, S. F. and Roberts, G. O. (2007). Convergence of heavy-tailed Monte Carlo Markov chain algorithms. Scand. J. Statist. 34, 781–815. Google Scholar
[16] Jarner, S. F. and Tweedie, R. L. (2003). Necessary conditions for geometric and polynomial ergodicity of random-walk-type Markov chains. Bernoulli 9, 559–578. Google Scholar
[17] 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. Google Scholar
[18] Kamatani, K. (2009). Metropolis-Hastings algorithms with acceptance ratios of nearly 1. Ann. Inst. Statist. Math. 61, 949–967. Google Scholar
[19] Kamatani, K. (2014). Efficient strategy for the Markov chain Monte Carlo in high-dimension with heavy-tailed target probability distribution. Preprint. Available at https://arxiv.org/abs/1412.6231v1. Google Scholar
[20] Kamatani, K. and Uchida, M. (2015). Hybrid multi-step estimators for stochastic differential equations based on sampled data. Statist. Inf. Stoch. Process. 18, 177–204. Google Scholar
[21] Law, K. J. H. (2014). Proposals which speed up function-space MCMC. J. Comput. Appl. Math. 262, 127–138. CrossRefGoogle Scholar
[22] Livingstone, S. (2015). Geometric ergodicity of the Random Walk Metropolis with position-dependent proposal covariance. Preprint. Available at https://arxiv.org/abs/1507.05780v2. Google Scholar
[23] Mengersen, K. L. and Tweedie, R. L. (1996). Rates of convergence of the Hastings and Metropolis algorithms. Ann. Statist. 24, 101–121. Google Scholar
[24] Meyn, S. and Tweedie, R. L. (2009). Markov Chains and Stochastic Stability, 2nd edn. Cambridge University Press>. Google Scholar
[25] Neal, R. M. (1999). Regression and classification using Gaussian process priors. In Bayesian Statistics, 6 (Alcoceber, 1998), Oxford University Press, New York, pp. 475–501. CrossRefGoogle Scholar
[26] Nummelin, E. (1984). General Irreducible Markov Chains and Nonnegative Operators (Camb. Tracts Math. 83). Cambridge University Press. CrossRefGoogle Scholar
[27] Resnick, S. I. (2008). Extreme Values, Regular Variation and Point Processes. Springer, New York. Google Scholar
[28] Roberts, G. O. and Tweedie, R. L. (1996). Geometric convergence and central limit theorems for multidimensional Hastings and Metropolis algorithms. Biometrika 83, 95–110. CrossRefGoogle Scholar
[29] Rudolf, D. and Sprungk, B. (2016). On a generalization of the preconditioned Crank–Nicolson Metropolis algorithm. Found. Comput. Math. 35pp. Google Scholar
[30] Tierney, L. (1994). Markov chains for exploring posterior distributions. Ann. Statist. 22, 1701–1762. Google Scholar
[31] Tuominen, P. and Tweedie, R. L. (1994). Subgeometric rates of convergence of f-ergodic Markov chains. Adv. Appl. Prob. 26, 775–798. Google Scholar