Efficient particle-based online smoothing in general hidden Markov models: the PaRIS algorithm

DOI10.3150/16-BEJ801zbMATH Open1392.62252arXiv1412.7550MaRDI QIDQ527474FDOQ527474

Authors: Jimmy Olsson, Johan Westerborn 

Publication date: 11 May 2017

Published in: Bernoulli (Search for Journal in Brave)

Abstract: This paper presents a novel algorithm, the particle-based, rapid incremental smoother (PaRIS), for efficient online approximation of smoothed expectations of additive state functionals in general hidden Markov models. The algorithm, which has a linear computational complexity under weak assumptions and very limited memory requirements, is furnished with a number of convergence results, including a central limit theorem. An interesting feature of PaRIS, which samples on-the-fly from the retrospective dynamics induced by the particle filter, is that it requires two or more backward draws per particle in order to cope with degeneracy of the sampled trajectories and to stay numerically stable in the long run with an asymptotic variance that grows only linearly with time.

Full work available at URL: https://arxiv.org/abs/1412.7550

Recommendations

• Sequential Monte Carlo smoothing for general state space hidden Markov models
• On-Line Inference for Hidden Markov Models via Particle Filters
• Particle-based online estimation of tangent filters with application to parameter estimation in nonlinear state-space models
• Particle smoothing via Markov chain Monte Carlo in general state space models
• Recursive computation of smoothed functionals of hidden Markovian processes using a particle approximation

zbMATH Keywords

sequential Monte Carlosmoothingcentral limit theoremonline estimationparticle filtergeneral hidden Markov modelsHoeffding-type inequalityparticle path degeneracy

Mathematics Subject Classification ID

Computational methods in Markov chains (60J22) Markov processes: estimation; hidden Markov models (62M05) Central limit and other weak theorems (60F05)

Cited In (18)

• Foreword. On sequential Monte Carlo: an overview
• Particle-based, rapid incremental smoother meets particle Gibbs
• On particle methods for parameter estimation in state-space models
• On Large Lag Smoothing for Hidden Markov Models
• Fast and Numerically Stable Particle-Based Online Additive Smoothing: The AdaSmooth Algorithm
• Adaptive online variance estimation in particle filters: the ALVar estimator
• On the two-filter approximations of marginal smoothing distributions in general state-space models
• Backward Importance Sampling for Online Estimation of State Space Models
• On backward smoothing algorithms
• A pseudo-marginal sequential Monte Carlo online smoothing algorithm
• Score-Based Parameter Estimation for a Class of Continuous-Time State Space Models
• Limits of accuracy for parameter estimation and localization in single-molecule microscopy via sequential Monte Carlo methods
• Particle-based online estimation of tangent filters with application to parameter estimation in nonlinear state-space models
• Online Smoothing for Diffusion Processes Observed with Noise
• On-Line Inference for Hidden Markov Models via Particle Filters
• Smoothing with couplings of conditional particle filters
• Variance estimation for sequential Monte Carlo algorithms: a backward sampling approach
• Linear prediction error methods for stochastic nonlinear models

This page was built for publication: Efficient particle-based online smoothing in general hidden Markov models: the PaRIS algorithm

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q527474)