Abstract
This paper is concerned with parameter estimation in linear and non-linear Itô type stochastic differential equations using Markov chain Monte Carlo (MCMC) methods. The MCMC methods studied in this paper are the Metropolis–Hastings and Hamiltonian Monte Carlo (HMC) algorithms. In these kind of models, the computation of the energy function gradient needed by HMC and gradient based optimization methods is non-trivial, and here we show how the gradient can be computed with a linear or non-linear Kalman filter-like recursion. We shall also show how in the linear case the differential equations in the gradient recursion equations can be solved using the matrix fraction decomposition. Numerical results for simulated examples are presented and discussed in detail.
Similar content being viewed by others
Notes
Originally called hybrid Monte Carlo (see, Neal 2011).
The Matlab codes can be obtained from the corresponding author on request.
References
Aït-Sahalia Y (2002) Maximum likelihood estimation of discretely sampled diffusions: a closed-form approximation approach. Econometrica 70(1):223–262
Aït-Sahalia Y (2008) Closed-form expansion for multivariate diffusions. Ann Stat 36(2):906–937
Andrieu C, Thoms J (2008) A tutorial on adaptive MCMC. Stat Comput 18(4):343–373
Andrieu C, Doucet A, Holenstein R (2010) Particle Markov chain Monte Carlo methods. R Stat Soc Ser B Stat Methodol 72(3):269–342
Berg EVD, Heemink AW, Lin HX, Schoenmakers J (2006) Probability density estimation in stochastic environmental models using reverse representations. Stoch Environ Res Risk Assess 20(1–2):126–139
Berger JO (1985) Statistical decision theory and Bayesian analysis. Springer, Berlin
Beskos A, Papaspiliopoulos O, Roberts G, Fearnhead P (2006) Exact and computationally efficient likelihood-based estimation for discretely observed diffusion processes (with discussion). J R Stat Soc Ser B Stat Methodol 68(3):333–382
Beskos A, Papaspiliopoulos O, Roberts G (2009) Monte Carlo maximum likelihood estimation for discretely observed diffusion processes. Ann Stat 37(1):223–245
Beskos A, Pillai NS, Roberts GO, Sanz-Serna JM, Stuart AM (2010) Optimal tuning of the hybrid Monte Carlo algorithm (submitted)
Bishop CM (1996) Neural networks for pattern recognition. Oxford University, Oxford
Brandt M, Santa-Clara P (2002) Simulated likelihood estimation of diffusions with an application to exchange rate dynamics in incomplete markets. J Financ Econ 63(2):161–210
Butcher JC (2003) Numerical methods for ordinary differential equations. Wiley, New York
Cappé O, Moulines E, Rydén T (2005) Inference in Hidden Markov models. Springer, Berlin
Chen L, Qin Z, Liu JS (2001) Exploring hybrid Monte Carlo in Bayesian computation. In: Proceedings of ISBA
Doucet A, Godsill SJ, Andrieu C (2000) On sequential Monte Carlo sampling methods for Bayesian filtering. Stat Comput 10(3):197–208
Duane S, Kennedy AD, Pendleton BJ, Roweth D (1987) Hybrid Monte Carlo. Phys Lett B 195(2):216–222
Elerian O (1998) A note on the existence of a closed form conditional transition density for the Milstein scheme. Working paper, Nuffield College, Oxford University
Elerian O, Chib S, Shephard N (2001) Likelihood inference for discretely observed non-linear diffusions. Econometrica 69(4):959–993
Eraker B (2001) MCMC analysis of diffusion models with application to finance. J Bus Econ Stat 19(2):177–191
Gelman A, Carlin JB, Stern HS, Rubin DB (2004) Bayesian data analysis. CRC/Chapman and Hall, London
Geyer CJ (1992) Practical Markov chain Monte Carlo. Stat Sci 7(4):473–483
Gilks WR, Richardson S, Spiegelhalter DJ (1996) Markov chain Monte Carlo in practice. Chapman and Hall/CRC, London
Golightly A, Wilkinson DJ (2006) Bayesian sequential inference for nonlinear multivariate diffusions. Stat Comput 16(4):323–338
Golightly A, Wilkinson DJ (2008) Bayesian inference for nonlinear multivariate diffusion models observed with error. Comput Stat Data Anal 52(3):1674–1693
Grewal MS, Andrews AP (2001) Kalman filtering, theory and practice using MATLAB. Wiley-IEEE Press, New York
Holder T, Leimkuhler B, Reich S (2001) Explicit variable step-size and time-reversible integration. Appl Numer Math 39(3–4):367–377
Hurn AS, Lindsay KA (1999) Estimating the parameters of stochastic differential equations. Mathe Comput Simul 48(4–6):373–384
Hurn AS, Lindsay KA, Martin VL (2003) On the efficacy of simulated maximum likelihood for estimating the parameters of stochastic differential equations. J Time Ser Anal 24(1):45–63
Jazwinski AH (1970) Stochastic processes and filtering theory. Academic Press, Edinburgh
Jeisman J (2005) Estimation of the Parameters of stochastic differential equations. Doctoral dissertation, Queensland University of Technology
Jensen B, Poulsen R (2002) Transition densities of diffusion processes: numerical comparison of approximation techniques. J Deriv 9(4):18–32
Jones CS (1998) A simple Bayesian method for the analysis of diffusion processes. Working paper, University of Pennsylvania
Julier SJ, Uhlmann JK (2004) Unscented filtering and nonlinear estimation. Proc IEEE 92(3):401–422
Karatzas I, Shreve SE (1991) Brownian motion and stochastic calculus. Springer, Berlin
Kloeden PE, Platen E (1999) Numerical solution to stochastic differential equations. Springer, Berlin
Liang F, Liu C, Carroll RJ (2010) Advanced Markov chain Monte Carlo methods: learning from past samples. Wiley, Chichester
Najfeld I, Havel TF (1995) Derivatives of the matrix exponential and their computation. Adv Appl Math 16(3):321–375
Neal RM (2011) MCMC using Hamiltonian dynamics. In: Brooks S, Gelman A, Jones GL, Meng XL (eds) Handbook of Markov chain Monte Carlo, chap 5. Chapman& Hall/CRC, London
Nielsen JN, Madsen H, Young PC (2000) Parameter estimation in stochastic differential equations: an overview. SIAM J Numer Anal 24:83–94
Ninness B, Henriksen S (2010) Bayesian system identification via Markov chain Monte Carlo techniques. Automatica 46(1):40–51
Øksendal B (2003) Stochastic differential equations: an introduction with applications. Springer, Berlin
Ornstein LS, Uhlenbeck GE (1930) On the theory of the Brownian motion. Phys Rev 36(5):823–841
Pedersen AR (1995) A new approach to maximum likelihood estimation for stochastic differential equations based on discrete observations. Scand J Stat 22(1):55–71
Särkkä S (2006) Recursive Bayesian inference on stochastic differential equations. Doctoral dissertation, Helsinki University of Technology
Särkkä S (2007) On unscented Kalman filtering for state estimation of continuous-time nonlinear systems. IEEE Trans Autom Control 52(9):1631–1641
Särkkä S, Vehtari A, Lampinen J (2007) Rao-Blackwellized particle filter for multiple target tracking. Inf Fusion J 8(1):2–15
Shoji I, Ozaki T (1998) Estimation for nonlinear stochastic differential equations by a local linearization method. Stoch Anal Appl 16(4):733–752
Singer H (2002) Parameter estimation of nonlinear stochastic differential equations: simulated maximum likelihood versus extended Kalman filter and Itô-Taylor expansion. J Comput Graph Stat 11(4):972–995
Singer H (2004) A survey of estimation methods for stochastic differential equastions. In: Proceedings of 6th international conference on social science methodology, Amsterdam
Singer H (2011) Continuous-discrete state-space modeling of panel data with nonlinear filter algorithms. AStA Adv Stat Anal 95(4):375–413
Sørensen H (2004) Parametric inference for diffusion processes observed at discrete points in time: a survey. Int Stat Rev 72(3):337–354
Stramer O, Bognar M (2011) Bayesian inference for irreducible diffusion processes using the pseudo-marginal approach. Bayesian Anal 6(2):231–258
Stramer O, Bognar M, Schneider P (2010) Bayesian inference for discretely sampled Markov processes with closed-form likelihood expansions. J Financ Econ 8(4):450–480
Young P (1980) Parameter estimation for continuous-time models: a survey. Automatica 17(1):23–39
Acknowledgments
We would like to acknowledge Aki Vehtari, Antti Solonen, Jouni Hartikainen, and Arno Solin for their contributions. We thank the Department of Mathematics and Physics in Lappeenranta University of Technology, the Department of Biomedical Engineering and Computational Science in Aalto University and the Academy of Finland for their financial support.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Mbalawata, I.S., Särkkä, S. & Haario, H. Parameter estimation in stochastic differential equations with Markov chain Monte Carlo and non-linear Kalman filtering. Comput Stat 28, 1195–1223 (2013). https://doi.org/10.1007/s00180-012-0352-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00180-012-0352-y