Particle Markov chain Monte Carlo methods

By Christophe Andrieu, Arnaud Doucet, Roman Holenstein. 2010

2,191 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability, Probabilistic & Bayesian Methods

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

In Probabilistic & Bayesian Methods

Citations in the corpus

An Introduction to Probabilistic Programming (2018, methodology): “This SMC algorithm can additionally be used as a building-block for particle MCMC algorithms (Andrieu et al., 2010), which uses a single SMC sweep of L particles as a block proposal in a MetropolisHastings algorithm.”
A New Approach to Probabilistic Programming Inference (2014, methodology): “We propose a novel combination of ideas from probabilistic programming [3] and particle Markov chain Monte Carlo (PMCMC) [1] that yields a new scheme for exploring and characterizing the set of probable execution traces.”
Venture: a higher-order probabilistic programming platform with programmable inference (2014, methodology): “…Gibbs algorithm from (Wood et al., 2014) with 30 particles would be expressed as follows: [INFER (pgibbs default ordered 30 100)] Note that Metropolis-Hastings transitions can be more effective than pure conditional sequential Monte Carlo for handling global parameters (Andrieu et al., 2010).”
Riemann manifold Langevin and Hamiltonian Monte Carlo methods (2011, methodology): “It would be interesting to consider the use of the MMALA and RMHMC sampling as a part of the particle MCMC methodology (Andrieu et al., 2010) for this particular model.”
Bayesian Synthetic Likelihood (2017, methodology): “(2016) employ BSL and compare it with particle MCMC (Andrieu et al., 2010) for applications in ecology and epidemiology (see also Fasiolo and Wood (2016)). Everitt et al. (2016) use BSL for models that have an intractable normalising constant.”
Bayesian Inference and Learning in Gaussian Process State-Space Models with Particle MCMC (2013, neutral): “We discuss marginalization of f in Section 3.1 and present the inference algorithms Sections 3.2 and 3.3.”
Biips: Software for Bayesian Inference with Interacting Particle Systems (2014, methodology): “Recently, algorithms have been developed that propose to use SMC algorithms within a MCMC algorithm (Andrieu et al. 2010).”
Delayed Sampling and Automatic Rao-Blackwellization of Probabilistic Programs (2017, methodology): “This has motivated more sophisticated methods for parameter estimation such as particle Markov chain Monte Carlo methods [1], also applied to probabilistic programs [28].”
Neural Adaptive Sequential Monte Carlo (2015, methodology): “SMC is particularly well-suited for performing inference in non-linear dynamical models with hidden variables, since filtering naturally decomposes into a sequence, and in many such cases it is the state-of-the-art inference method [2, 3].”
The correlated pseudomarginal method (2015, methodology): “Finally, a sequential extension of the particle marginal MH algorithm (Andrieu et al., 2010), a PM method, has been proposed in (Chopin et al.”
State Space LSTM Models with Particle MCMC Inference (2017, methodology): “We present a general parameter estimation scheme for the proposed class of models based on sequential Monte Carlo (SMC) (Doucet et al., 2001), in particular the Particle Gibbs (Andrieu et al., 2010).”
Bayesian State-Space Modelling on High-Performance Hardware Using LibBi (2013, methodology): “Variants of sequential Monte Carlo (SMC, Doucet, Freitas, and Gordon 2001) are particularly attractive, including particle Markov chain Monte Carlo (PMCMC, Andrieu, Doucet, and Holenstein 2010) and SMC2 (Chopin, Jacob, and Papaspiliopoulos 2013).”
Dynamic Prediction Pools: An Investigation of Financial Frictions and Forecasting Performance (2016, positive): “In order to generate draws from the pseudo-posterior we use a particle-MCMC technique (see, Andrieu et al. (2010) as well as the online appendix) that combines a bootstrap particle filter with a random-walk Metropolis-Hastings algorithm.”
Sequential Monte Carlo with Highly Informative Observations (2014, methodology): “Unauthorized reproduction of this article is prohibited. particle marginal Metropolis–Hastings (PMMH, Andrieu, Coucet, and Holenstein [1]), then the CAR-based metric (4.3) is most appropriate.”
An efficient stochastic approximation EM algorithm using conditional particle filters (2013, methodology): “Recently, however, a framework for Bayesian inference referred to as particle Markov chain Monte Carlo (PMCMC) has been developed [7].”