Abstract
In this paper we develop set of novel Markov chain Monte Carlo algorithms for Bayesian smoothing of partially observed non-linear diffusion processes. The sampling algorithms developed herein use a deterministic approximation to the posterior distribution over paths as the proposal distribution for a mixture of an independence and a random walk sampler. The approximating distribution is sampled by simulating an optimized time-dependent linear diffusion process derived from the recently developed variational Gaussian process approximation method. The novel diffusion bridge proposal derived from the variational approximation allows the use of a flexible blocking strategy that further improves mixing, and thus the efficiency, of the sampling algorithms. The algorithms are tested on two diffusion processes: one with double-well potential drift and another with SINE drift. The new algorithm’s accuracy and efficiency is compared with state-of-the-art hybrid Monte Carlo based path sampling. It is shown that in practical, finite sample applications the algorithm is accurate except in the presence of large observation errors and low observation densities, which lead to a multi-modal structure in the posterior distribution over paths. More importantly, the variational approximation assisted sampling algorithm outperforms hybrid Monte Carlo in terms of computational efficiency, except when the diffusion process is densely observed with small errors in which case both algorithms are equally efficient.
Similar content being viewed by others
References
Alexander F, Eyink G, Restrepo J (2005) Accelerated Monte Carlo for optimal estimation of time series. J Stat Phys 119: 1331–1345
Andrieu C, de Freitas D, Doucet A, Jordan M (2003) An introduction to MCMC for machine learning. Mach Learn 50: 5–43
Andrieu C, Doucet A, Holenstein R (2010) Particle Markov Chain Monte Carlo methods. J R Statist Soc B 72: 1–33
Archambeau C, Cornford D, Opper M, Shawe-Tayler J (2007) Gaussian Process approximations of stochastic differential equations. J Mach Learn Res Workshop and Conference Proceedings 1: 1–16
Archambeau C, Opper M, Shen Y, Cornford D, Shawe-Tayler J (2008) Variational inference for diffusion processes. In: Platt C, Koller D, Singer Y, Roweis S (eds) Neural information processing systems (NIPS), vol 20. The MIT Press, Cambridge, pp 17–24
Beskos A, Papaspiliopoulous O, Roberts GO, Fearnhead P (2006) Exact and computationally efficient likelihood-based estimation for discretely observed diffusion processes. J R Statist Soc B 68: 333–382
Beskos A, Papaspiliopoulous O, Roberts GQ (2008) A factorisation of diffusion measure and finite sample path construction. Methodol Comput Appl Probab 10: 85–104
de Freitas N, H⌽jen-S⌽rensen P, Jordan M, Russell S (2001) Variational MCMC. In: Proceedings of the 17th annual conference on uncertainty in artificial intelligence. Morgan Kaufmann Publishers Inc., San Francisco, CA, pp 120–127
Derber J (1989) A variational continuous assimilation technique. Mon Wea Rev 117: 2437–2446
Duane S, Kennedy AD, Pendleton BJ, Roweth D (1987) Hybrid Monte Carlo. Phys Lett B 55: 2774–2777
Durham GB, Gallant AR (2002) Numerical techniques for maximum likelihood estimation of continuous-time diffusion process. J Bus Econom Stat 20: 297–338
Elerian O, Chib S, Shephard N (2001) Likelihood inference for discretely observed nonlinear diffusions. Econometrica 69: 959–993
Eraker B (2001) Markov Chain Monte Carlo analysis of diffusion models with application to finance. J Bus Econ Statist 19: 177–191
Evensen G (1994) Sequential data assimilation with a non-linear quasi-geostrophic model using Monte Carlo methods to forecast error statistics. J Geophys Res 99: 10,143–10,162
Evensen G (2000) An ensemble Kalman smoother for nonlinear dynamics. Mon Wea Rev 128: 1852–1867
Eyink GL, Restrepo JM, Alexander FJ (2004) A mean-field approximation in data assimilation for nonlinear dynamics. Physica D 194: 347–368
Golightly A, Wilkinson GJ (2006) Bayesian sequential inference for nonlinear multivariate diffusions. Stat Comput 16: 323–338
Golightly A, Wilkinson GJ (2008) Bayesian inference for nonlinear multivariate diffusion models observed with error. Comput Stat Data Anal 52: 1674–1693
Haario H, Laine M, Mira A, Saksman E (2006) Dram: efficient adaptive MCMC. Stat Comput 16: 339–354
Hastings WK (1970) Monte Carlo sampling methods using Markov chains and their applications. Biometrika 57: 97–109
Honerkamp J (1994) Stochastic dynamical systems. VCH, Weinheim
Jazwinski AH (1970) Stochastic processes and filtering theory. Academic Press, New York
Julier SJ, Uhlmann J, Durrant-Whyte H (2000) A new method for the nonlinear tranformation of means and covariances in filters and estimators. IEEE Trans Autom Control 45: 477–482
Kalman RE, Bucy R (1961) New results in linear filtering and prediction theory. J Basic Eng D 83: 95–108
Kalnay E (2003) Atmospheric modelling, data assimilation and predictability. Cambridge University Press, Cambridge
Kitagawa G (1987) Non-Gaussian state space modelling of non-stationary time series. J Am Stat Assoc 82: 503–514
Kitagawa G (1996) Monte Carlo filter and smoother for non-Gaussian nonlinear state space models. J Comput Graph Stat 5: 1–25
Klöden PE, Platen E (1992) Numerical solution of stochastic differential equations. Spinger, Berlin
Kushner HJ (1967) Dynamical equations for optimal filter. J Differ Equ 3: 179–190
Liu JS (2001) Monte Carlo strategies in scientific computing. Spinger, Berlin
Miller RN, Carter EF, Blue ST (1999) Data assimilation into nonlinear stochastic models. Tellus A 51: 167–194
Mira A (2001) On Metropolis-Hastings algorithms with delayed rejection. Metron LIX: 231–241
Ozaki T (1992) A bridge between nonlinear time series models and nonlinear stochastic dynamical systems: a local linearization approach. Stat Sinica 2: 113–135
Papaspiliopolous O, Roberts GO, Skold M (2003) Non-centered parameterisations for hierarchical models and data augmentation. In: Bayesian Statistics 7, pp 307–326
Pardoux E (1982) équations du filtrage non linéaire de la prédiction et du lissage. Stochastics 6: 193–231
Rabier F, Jarvinen H, Klinker E, Mahfouf JF, Simmons A (2000) The ecmwf operational implementation of four-dimensional variational assimilation. Part i: experimental results with simplified physics. Quart J Roy Met Soc 126: 1143–1170
Roberts GQ, Stramer O (2001) On inferencee for partially observed non-linear diffusion models using Metropolis-Hasting algorithm. Biometrika 88: 603–621
Stuart AM, Voss J, Winberg P (2004) Conditional path sampling of SDEs and the Langevin MCMC method. Commun Math Sci 2: 685–697
Wan E, van der Merwe R (2001) The unscented Kalman filter. In: Haykin S (ed) Kalman filtering and neural networks. Wiley, New York, pp 207–219
Wilkinson D (2006) Stochastic modelling for systems biology. Chapman & Hall/CRC, Boca Raton
Wilkinson DJ, Golightly A (2010) Markov Chain Monte Carlo algorithms for SDE parameter estimation. In: Learning and inference in computational systems biology, pp 253–276
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Shen, Y., Cornford, D., Opper, M. et al. Variational Markov chain Monte Carlo for Bayesian smoothing of non-linear diffusions. Comput Stat 27, 149–176 (2012). https://doi.org/10.1007/s00180-011-0246-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00180-011-0246-4