On Large Lag Smoothing for Hidden Markov Models

DOI10.1137/18M1198004zbMATH Open1434.62175arXiv1804.07117OpenAlexW2992092112WikidataQ126620637 ScholiaQ126620637MaRDI QIDQ5203791FDOQ5203791

Authors: Jeremie Houssineau, Ajay Jasra, Sumeetpal S. Singh 

Publication date: 9 December 2019

Published in: SIAM Journal on Numerical Analysis (Search for Journal in Brave)

Abstract: In this article we consider the smoothing problem for hidden Markov models (HMM). Given a hidden Markov chain ${X_n}_{ngeq0}$ and observations ${Y_n}_{ngeq0}$, our objective is to compute $mathbbE[varphi(X_0,dots,X_k)|y_0,dots,y_n]$ for some real-valued, integrable functional $varphi$ and $k$ fixed, $klln$ and for some realisation $(y_0,dots,y_n)$ of $(Y_0,dots,Y_n)$. We introduce a novel application of the multilevel Monte Carlo (MLMC) method with a coupling based on the Knothe-Rosenblatt rearrangement. We prove that this method can approximate the afore-mentioned quantity with a mean square error (MSE) of $mathcalO(epsilo{n}^2)$, for arbitrary $epsilon>0$ with a cost of $mathcalO(epsilo{n}^{-2})$. This is in contrast to the same direct Monte Carlo method, which requires a cost of $mathcalO(nepsilo{n}^{-2})$ for the same MSE. The approach we suggest is, in general, not possible to implement, so the optimal transport methodology of cite{span} is used, which directly approximates our strategy. We show that our theoretical improvements are achieved, even under approximation, in several numerical examples.

Full work available at URL: https://arxiv.org/abs/1804.07117

Recommendations

• On approximation of smoothing probabilities for hidden Markov models
• Recursive smoothers for hidden discrete-time Markov chains
• A correction to ``On approximation of smoothing probabilities for hidden Markov models
• Exact smoothing in hidden conditionally Markov switching linear models
• Consistent Estimation of the Filtering and Marginal Smoothing Distributions in Nonparametric Hidden Markov Models
• Likelihood computation for hidden Markov models via generalized two-filter smoothing
• Sequential Monte Carlo smoothing for general state space hidden Markov models
• Reduced-Complexity Estimation for Large-Scale Hidden Markov Models
• Risk-sensitive filtering and smoothing for hidden Markov models
• scientific article; zbMATH DE number 2144803

zbMATH Keywords

smoothingoptimal transportmultilevel Monte Carlo

Mathematics Subject Classification ID

Markov processes: estimation; hidden Markov models (62M05) Monte Carlo methods (65C05) Markov chains (discrete-time Markov processes on discrete state spaces) (60J10) Approximations to statistical distributions (nonasymptotic) (62E17)

Cites Work

• Inference in hidden Markov models.
• Title not available (Why is that?)
• Particle Markov Chain Monte Carlo Methods
• Title not available (Why is that?)
• On particle methods for parameter estimation in state-space models
• Chebyshev and Fourier spectral methods.
• Particle Gibbs with ancestor sampling
• Multilevel Monte Carlo Path Simulation
• A backward particle interpretation of Feynman-Kac formulae
• Efficient particle-based online smoothing in general hidden Markov models: the PaRIS algorithm
• Title not available (Why is that?)
• On particle Gibbs sampling
• Multilevel Monte Carlo methods
• On the Kantorovich-Rubinstein theorem
• Unbiased estimation with square root convergence for SDE models
• Multilevel Particle Filters
• Multilevel ensemble Kalman filtering
• Multilevel ensemble transform particle filtering
• Smoothing with couplings of conditional particle filters
• Inference via low-dimensional couplings
• A multiscale strategy for Bayesian inference using transport maps
• On the behaviour of the backward interpretation of Feynman-Kac formulae under verifiable conditions
• Gibbs Flow for Approximate Transport with Applications to Bayesian Computation

Cited In (2)

• Recursive smoothers for hidden discrete-time Markov chains
• A correction to ``On approximation of smoothing probabilities for hidden Markov models

This page was built for publication: On Large Lag Smoothing for Hidden Markov Models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5203791)