Abstract
Riemann manifold Hamiltonian Monte Carlo (RMHMC) has the potential to produce high-quality Markov chain Monte Carlo output even for very challenging target distributions. To this end, a symmetric positive definite scaling matrix for RMHMC is proposed. The scaling matrix is obtained by applying a modified Cholesky factorization to the potentially indefinite negative Hessian of the target log-density. The methodology is able to exploit the sparsity of the Hessian, stemming from conditional independence modeling assumptions, and thus admit fast implementation of RMHMC even for high-dimensional target distributions. Moreover, the methodology can exploit log-concave conditional target densities, often encountered in Bayesian hierarchical models, for faster sampling and more straightforward tuning. The proposed methodology is compared to alternatives for some challenging targets and is illustrated by applying a state-space model to real data.
Similar content being viewed by others
References
Aït-Sahalia, Y.: Testing continuous-time models of the spot interest rate. Rev. Financ. Stud. 9(2), 385–426 (1996)
Andrieu, C., Doucet, A., Holenstein, R.: Particle Markov chain Monte Carlo methods. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 72(3), 269–342 (2010)
Beskos, A., Pillai, N., Roberts, G., Sanz-Serna, J.-M., Stuart, A.: Optimal tuning of the hybrid Monte Carlo algorithm. Bernoulli 19(5A), 1501–1534 (2013)
Betancourt, M.: A general metric for Riemannian manifold Hamiltonian Monte Carlo. In: Nielsen, F., Barbaresco, F. (eds.) Geometric Science of Information, Volume 8085 of Lecture Notes in Computer Science, vol. 8085, pp. 327–334. Springer, Berlin (2013a)
Betancourt, M.: Identifying the Optimal Integration Time in Hamiltonian Monte Carlo. arXiv:1601.00225 (2016)
Betancourt, M., Byrne, S., Livingstone, S., Girolami, M.: The Geometric Foundations of Hamiltonian Monte Carlo. Bernoulli 23(4A), 2257–2298 (2017). doi:10.3150/16-BEJ810
Betancourt, M.J.: Generalizing the No-u-turn Sampler to Riemannian Manifolds. arXiv:1304.1920 (2013b)
Betancourt, M.J., Girolami, M.: Hamiltonian Monte Carlo for Hierarchical Models. arXiv:1312.0906 (2013)
Calderhead, B.: A general construction for parallelizing Metropolis–Hastings algorithms. Proc Natl Acad Sci 111(49), 17408–17413 (2014)
Carpenter, B., Gelman, A., Hoffman, M., Lee, D., Goodrich, B., Betancourt, M., Brubaker, M., Guo, J., Li, P., Riddell, A.: Stan: a probabilistic programming language. J. Stat. Softw. 76(1), 1–32 (2017)
Chan, K.C., Karolyi, G.A., Longstaff, F.A., Sanders, A.B.: An empirical comparison of alternative models of the short-term interest rate. J. Finance 47(3), 1209–1227 (1992)
Davis, T.A.: Direct methods for sparse linear systems. In: Fundamentals of Algorithms 2. Society for Industrial and Applied Mathematics, Philadelphia, PA, USA (2006)
Duane, S., Kennedy, A., Pendleton, B.J., Roweth, D.: Hybrid Monte Carlo. Phys. Lett. B 195(2), 216–222 (1987)
Flury, T., Shephard, N.: Bayesian inference based only on simulated likelihood: particle filter analysis of dynamic economic models. Econom. Theory 27(Special Issue 05), 933–956 (2011)
Gelman, A., Carlin, J.B., Stern, H.S., Dunson, D.B., Vehtari, A., Rubin, D.: Bayesian Data Analysis, 3rd edn. CRC Press, Boca Raton (2014)
Geweke, J., Tanizaki, H.: On Markov chain Monte Carlo methods for nonlinear and non-gaussian state-space models. Commun. Stat. Simul. Comput. 28(4), 867–894 (1999)
Geweke, J., Tanizaki, H.: Note on the sampling distribution for the Metropolis–Hastings algorithm. Commun. Stat. Theory Methods 32(4), 775–789 (2003)
Geyer, C.J.: Practical Markov chain Monte Carlo. Stat. Sci. 7(4), 473–483 (1992)
Gill, P., Murray, W.: Newton-type methods for unconstrained and linearly constrained optimization. Math. Program. 7, 311–350 (1974)
Gill, P.E., Murray, W., Wright, M.H.: Practical Optimization. Academic Press, London (1981)
Girolami, M., Calderhead, B.: Riemann manifold Langevin and Hamiltonian Monte Carlo methods. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 73(2), 123–214 (2011)
Golub, G.H., van Loan, C.F.: Matrix Computations, 3rd edn. The Johns Hopkins University Press, Baltimore (1996)
Griewank, A.: Evaluating Derivatives: Principles and Techniques of Algorithmic Differentiation. SIAM, Philadelphia (2000)
Grothe, O., Kleppe, T.S., Liesenfeld, R.: The Gibbs Sampler with Particle Efficient Importance Sampling for State-space Models. arXiv:1601.01125 (2016)
Guerrera, T., Rue, H., Simpson, D.: Discussion of Riemann manifold Langevin and Hamiltonian Monte Carlo by Girolami and Calderhead. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 73(2), 123–214 (2011)
Haario, H., Saksman, E., Tamminen, J.: Adaptive proposal distribution for random walk Metropolis algorithm. Comput. Stat. 14(3), 375–395 (1999)
Hoffman, M.D., Gelman, A.: The no-u-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo. J. Mach. Learn. Res. 15, 1593–1623 (2014)
Hogan, R .J.: Fast reverse-mode automatic differentiation using expression templates in C++. ACM Trans. Math. Softw. 40(4), 26:1–26:16 (2014)
Jasra, A., Singh, S.: Discussion of Riemann manifold Langevin and Hamiltonian Monte Carlo by Girolami and Calderhead. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 73(2), 123–214 (2011)
Kleppe, T.S.: Adaptive step size selection for Hessian-based manifold Langevin samplers. Scand. J. Stat. 43(3), 788–805 (2016)
Kristensen, K., Nielsen, A., Berg, C., Skaug, H., Bell, B.: Tmb: automatic differentiation and Laplace approximation. J. Stat. Softw. 70(1), 1–21 (2016)
Lan, S., Stathopoulos, V., Shahbaba, B., Girolami, M.: Markov chain Monte Carlo from Lagrangian dynamics. J. Comput. Graph. Stat. 24(2), 357–378 (2015)
Leimkuhler, B., Reich, S.: Simulating Hamiltonian dynamics. Cambridge University Press, Cambridge (2004)
Lindgren, F., Rue, H., Lindström, J.: An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach. J. R. Stat. Soc. Ser. B 73(4), 423–498 (2011)
Lindsten, F., Jordan, M.I., Schön, T.B.: Particle Gibbs with ancestor sampling. J. Mach. Learn. Res. 15, 2145–2184 (2014)
Liu, J.S.: Monte Carlo Strategies in Scientific Computing. Springer Series in Statistics. Springer, Berlin (2001)
Mannseth, J., Kleppe, T.S., Skaug, H.J.: On the application of improved symplectic integrators in Hamiltonian Monte Carlo. Commun. Stat. Simul. Comput. 1–10 (2017). doi:10.1080/03610918.2017.1283703
Martin, J., Wilcox, L., Burstedde, C., Ghattas, O.: A stochastic Newton MCMC method for large-scale statistical inverse problems with application to seismic inversion. SIAM J. Sci. Comput. 34(3), A1460–A1487 (2012)
Neal, R.M.: Bayesian Learning for Neural Networks. Number 118 in Lecture Notes in Statistics. Springer, Berlin (1996)
Neal, R.M.: Slice sampling. Ann. Stat. 31(3), 705–767 (2003)
Neal, R. M.: MCMC using Hamiltonian dynamics. In: Brooks, S., Gelman, A., Jones, G. & Meng, X.-L. (eds) Handbook of Markov chain Monte Carlo, pp. 113–162. Chapman & Hall/CRC, Boca Raton (2010)
Nocedal, J., Wright, S.J.: Numerical Optimization. Springer, Berlin (1999)
Pitt, M.K., dos Santos Silva, R., Giordani, P., Kohn, R.: On some properties of Markov chain Monte Carlo simulation methods based on the particle filter. J. Econom. 171(2), 134–151 (2012)
Qi, Y., and Minka, T.P.:Hessian-based Markov Chain Monte-Carlo algorithms. Unpublished manuscript (2002). https://www.cs.purdue.edu/homes/alanqi/papers/qi-minka-HMH-AMIT-02.pdf
Robert, C.P., Casella, G.: Monte Carlo Statistical Methods, 2nd edn. Springer, Berlin (2004)
Roberts, G., Stramer, O.: Langevin diffusions and Metropolis–Hastings algorithms. Methodol. Comput. Appl. Probab. 4(4), 337–357 (2002)
Rue, H.: Fast sampling of Gaussian Markov random fields. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 63(2), 325–338 (2001)
Rue, H., Held, L.: Gaussian Markov Random Fields: Theory and Application. Chapman and Hall-CRC Press, Boca Raton (2005)
Rue, H., Martino, S., Chopin, N.: Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 71(2), 319–392 (2009)
Sanz-Serna, J.M.: Discussion of Riemann manifold Langevin and Hamiltonian Monte Carlo by Girolami and Calderhead. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 73(2), 123–214 (2011)
Scharth, M., Kohn, R.: Particle efficient importance sampling. J. Econom. 190(1), 133–147 (2016)
Schnabel, R., Eskow, E.: A new modified Cholesky factorization. SIAM J. Sci. Stat. Comput. 11(6), 1136–1158 (1990)
Schnabel, R., Eskow, E.: A revised modified Cholesky factorization algorithm. SIAM J. Optim. 9(4), 1135–1148 (1999)
Shephard, N., Pitt, M.K.: Likelihood analysis of non-Gaussian measurement time series. Biometrika 84, 653–667 (1997)
Skaug, H., Fournier, D.: Automatic approximation of the marginal likelihood in non-Gaussian hierarchical models. Comput. Stat. Data Anal. 56, 699–709 (2006)
Xifara, T., Sherlock, C., Livingstone, S., Byrne, S., Girolami, M.: Langevin diffusions and the Metropolis-adjusted Langevin algorithm. Stat. Probab. Lett. 91, 14–19 (2014)
Author information
Authors and Affiliations
Corresponding author
Additional information
The author would like to thank the Editor, the Associate Editor, two referees, Michael Betancourt, Hans J. Skaug and Anders Tranberg for comments that have sparked many improvements.
Rights and permissions
About this article
Cite this article
Kleppe, T.S. Modified Cholesky Riemann Manifold Hamiltonian Monte Carlo: exploiting sparsity for fast sampling of high-dimensional targets. Stat Comput 28, 795–817 (2018). https://doi.org/10.1007/s11222-017-9763-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11222-017-9763-5