On disturbance state-space models and the particle marginal Metropolis-Hastings sampler

DOI10.1137/130915376MaRDI QIDQ5397886zbMATH OpenOpenAlexFDO

Authors Lawrence M. Murray, Emlyn M. Jones, John Parslow

Publication date 25 February 2014

Published in SIAM/ASA Journal on Uncertainty Quantification (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1202.6159

zbMATH Keywords

Bayesian statisticsparticle filterparticle Markov chain Monte Carlosequential Monte Carlostate-space modelunscented Kalman filter

Mathematics Subject Classification ID

Monte Carlo methods (65C05)

Abstract: We investigate nonlinear state-space models without a closed-form transition density, and propose reformulating such models over their latent noise variables rather than their latent state variables. In doing so the tractable noise density emerges in place of the intractable transition density. For importance sampling methods such as the auxiliary particle filter, this enables importance weights to be computed where they could not be otherwise. As case studies we take two multivariate marine biogeochemical models and perform state and parameter estimation using the particle marginal Metropolis-Hastings sampler. For the particle filter within this sampler, we compare several proposal strategies over noise variables, all based on lookaheads with the unscented Kalman filter. These strategies are compared using conventional means for assessing Metropolis-Hastings efficiency, as well as with a novel metric called the conditional acceptance rate for assessing the consequences of using an estimated, and not exact, likelihood. Results indicate the utility of reformulating the model over noise variables, particularly for fast-mixing process models.

Recommendations

• A flexible particle Markov chain Monte Carlo method
• Particle Metropolis-Hastings using gradient and Hessian information
• Bayesian estimation of state-space models using the Metropolis-Hastings algorithm within Gibbs sampling.
• The Gibbs sampler with particle efficient importance sampling for state-space models
• Monte Carlo Estimation for Nonlinear Non-Gaussian State Space Models

Cited in

(10)

• Conditional particle filters with diffuse initial distributions
• Gradient free parameter estimation for hidden Markov models with intractable likelihoods
• Sequential Bayesian inference for implicit hidden Markov models and current limitations
• Augmentation schemes for particle MCMC
• Sequential Monte Carlo with Highly Informative Observations
• Bayesian inference for nonlinear structural time series models
• Path storage in the particle filter
• Approximate Bayesian Computation for a Class of Time Series Models
• The Block-Correlated Pseudo Marginal Sampler for State Space Models
• Improved auxiliary particle filtering with a fully deterministic resampling scheme to treat posterior tails accurately

This page was built for publication: On disturbance state-space models and the particle marginal Metropolis-Hastings sampler

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5397886)