Sequential Monte Carlo smoothing for general state space hidden Markov models

DOI10.1214/10-AAP735MaRDI QIDQ657691zbMATH OpenOpenAlexFDO

Authors Randal Douc, Aurélien Garivier, Jimmy Olsson, Eric Moulines

Publication date 10 January 2012

Published in The Annals of Applied Probability (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1202.2945

zbMATH Keywords

smoothingparticle filterhidden Markov modelssequential Monte Carlo methods

Mathematics Subject Classification ID

Interacting random processes; statistical mechanics type models; percolation theory (60K35) Stationary stochastic processes (60G10) Self-similar stochastic processes (60G18)

Abstract: Computing smoothing distributions, the distributions of one or more states conditional on past, present, and future observations is a recurring problem when operating on general hidden Markov models. The aim of this paper is to provide a foundation of particle-based approximation of such distributions and to analyze, in a common unifying framework, different schemes producing such approximations. In this setting, general convergence results, including exponential deviation inequalities and central limit theorems, are established. In particular, time uniform bounds on the marginal smoothing error are obtained under appropriate mixing conditions on the transition kernel of the latent chain. In addition, we propose an algorithm approximating the joint smoothing distribution at a cost that grows only linearly with the number of particles.

Recommendations

• Sequential Monte Carlo smoothing with application to parameter estimation in nonlinear state space models
• Smoothing algorithms for state-space models
• Monte Carlo Smoothing for Nonlinear Time Series
• Non-asymptotic deviation inequalities for smoothed additive functionals in nonlinear state-space models
• Central limit theorem for sequential Monte Carlo methods and its application to Bayesian inference

Cites work

• scientific article; zbMATH DE number 3954145 (Why is no real title available?)
• scientific article; zbMATH DE number 46153 (Why is no real title available?)
• scientific article; zbMATH DE number 2061746 (Why is no real title available?)
• scientific article; zbMATH DE number 1500585 (Why is no real title available?)
• scientific article; zbMATH DE number 2106098 (Why is no real title available?)
• A backward particle interpretation of Feynman-Kac formulae
• A note on auxiliary particle filters
• A sequential smoothing algorithm with linear computational cost
• Array algorithms for H/sup ∞/ estimation
• Central limit theorem for sequential Monte Carlo methods and its application to Bayesian inference
• Discrete time nonlinear filters with informative observations are stable
• Filtering via Simulation: Auxiliary Particle Filters
• Forgetting the initial distribution for hidden Markov models
• Inference in hidden Markov models.
• Intrinsic methods in filter stability
• Limit theorems for weighted samples with applications to sequential Monte Carlo methods
• Model robustness of finite state nonlinear filtering over the infinite time horizon
• Monte Carlo Smoothing for Nonlinear Time Series
• Monte Carlo strategies in scientific computing
• Monte Carlo techniques to estimate the conditional expectation in multi-stage non-linear filtering†
• On Approximate Maximum-Likelihood Methods for Blind Identification: How to Cope With the Curse of Dimensionality
• On discrete time ergodic filters with wrong initial data
• On the auxiliary particle filter
• Particle methods: an introduction with applications
• Recursive Monte Carlo filters: algorithms and theoretical analysis
• Sequential Monte Carlo Methods in Practice
• Smoothing algorithms for state-space models
• Stability of nonlinear filters in nonmixing case
• The simulation smoother for time series models
• The stability of conditional Markov processes and Markov chains in random environments
• Uniform observability of hidden Markov models and filter stability for unstable signals
• Uniform time average consistency of Monte Carlo particle filters

Cited in

(46)

• Coupled conditional backward sampling particle filter
• On the use of sequential Monte Carlo methods for approximating smoothing functionals, with application to fixed parameter estimation
• The HESSIAN method: highly efficient simulation smoothing, in a nutshell
• Linear prediction error methods for stochastic nonlinear models
• Particle-based online estimation of tangent filters with application to parameter estimation in nonlinear state-space models
• Fast and Numerically Stable Particle-Based Online Additive Smoothing: The AdaSmooth Algorithm
• Particle-kernel estimation of the filter density in state-space models
• Sequential Monte Carlo sampling in hidden Markov models of nonlinear dynamical systems
• Backward Importance Sampling for Online Estimation of State Space Models
• On backward smoothing algorithms
• Approximate Bayesian Computation for Smoothing
• Non-asymptotic deviation inequalities for smoothed additive functionals in nonlinear state-space models
• Exact smoothing in hidden conditionally Markov switching linear models
• Monte Carlo Smoothing for Nonlinear Time Series
• Direct, prediction- and smoothing-based Kalman and particle filter algorithms
• Sampling latent states for high-dimensional non-linear state space models with the embedded HMM method
• On the behaviour of the backward interpretation of Feynman-Kac formulae under verifiable conditions
• On some extensions of the sequential Monte Carlo methods in high-order hidden Markov models
• A method for high-dimensional smoothing
• Long-term stability of sequential Monte Carlo methods under verifiable conditions
• Online Smoothing for Diffusion Processes Observed with Noise
• Calibrating the exponential Ornstein-Uhlenbeck multiscale stochastic volatility model
• Estimation in the partially observed stochastic Morris-Lecar neuronal model with particle filter and stochastic approximation methods
• A new particle filter based on smooth variable structure filter
• Convergence of sequential quasi-Monte Carlo smoothing algorithms
• A sharp first order analysis of Feynman-Kac particle models. I: Propagation of chaos
• Likelihood computation for hidden Markov models via generalized two-filter smoothing
• On particle methods for parameter estimation in state-space models
• Stability properties of some particle filters
• On robust input design for nonlinear dynamical models
• Efficient particle-based online smoothing in general hidden Markov models: the PaRIS algorithm
• On the two-filter approximations of marginal smoothing distributions in general state-space models
• Smoothing with couplings of conditional particle filters
• A backward particle interpretation of Feynman-Kac formulae
• MCMC for state-space models
• A tutorial on particle filters
• Sequential Monte Carlo smoothing with application to parameter estimation in nonlinear state space models
• Particle filtering for continuous-time hidden Markov models
• On approximation of smoothing probabilities for hidden Markov models
• Recursive computation of smoothed functionals of hidden Markovian processes using a particle approximation
• Sequential Monte Carlo methods
• A pseudo-marginal sequential Monte Carlo online smoothing algorithm
• On Large Lag Smoothing for Hidden Markov Models
• Smoothing distributions for conditional Fleming-Viot and Dawson-Watanabe diffusions
• Variance estimation for sequential Monte Carlo algorithms: a backward sampling approach
• Uniform Stability of a Particle Approximation of the Optimal Filter Derivative

This page was built for publication: Sequential Monte Carlo smoothing for general state space hidden Markov models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q657691)