Home > Research > Publications & Outputs > Exact Bayesian inference for discretely observe...

Research

Associated organisational units

Electronic data

  • MJPexactJCGSfinal

    Accepted author manuscript, 472 KB, PDF document

    Available under license: CC BY: Creative Commons Attribution 4.0 International License

Links

Text available via DOI:

Keywords

View graph of relations

Exact Bayesian inference for discretely observed Markov Jump Processes using finite rate matrices

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

Exact Bayesian inference for discretely observed Markov Jump Processes using finite rate matrices. / Sherlock, Chris; Golightly, Andrew.
In: Journal of Computational and Graphical Statistics, Vol. 32, No. 1, 31.01.2023, p. 36-48.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Sherlock, C & Golightly, A 2023, 'Exact Bayesian inference for discretely observed Markov Jump Processes using finite rate matrices', Journal of Computational and Graphical Statistics, vol. 32, no. 1, pp. 36-48. https://doi.org/10.1080/10618600.2022.2093886

APA

Sherlock, C., & Golightly, A. (2023). Exact Bayesian inference for discretely observed Markov Jump Processes using finite rate matrices. Journal of Computational and Graphical Statistics, 32(1), 36-48. https://doi.org/10.1080/10618600.2022.2093886

Vancouver

Sherlock C, Golightly A. Exact Bayesian inference for discretely observed Markov Jump Processes using finite rate matrices. Journal of Computational and Graphical Statistics. 2023 Jan 31;32(1):36-48. Epub 2022 Jun 29. doi: 10.1080/10618600.2022.2093886

Author

Sherlock, Chris ; Golightly, Andrew. / Exact Bayesian inference for discretely observed Markov Jump Processes using finite rate matrices. In: Journal of Computational and Graphical Statistics. 2023 ; Vol. 32, No. 1. pp. 36-48.

Bibtex

@article{804f1075e3de4944b72a87fb352dbc33,
title = "Exact Bayesian inference for discretely observed Markov Jump Processes using finite rate matrices",
abstract = "We present new methodologies for Bayesian inference on the rate parameters of a discretely observed continuous-time Markov jump process with a countably infinite statespace. The usual method of choice for inference, particle Markov chain Monte Carlo (particle MCMC), struggles when the observation noise is small. We consider the most challenging regime of exact observations and provide two new methodologies for inference in this case: the minimal extended statespace algorithm (MESA) and the nearly minimal extended statespace algorithm (nMESA). By extending the Markov chain Monte Carlo statespace, both MESA and nMESA use the exponentiation of finite rate matrices to perform exact Bayesian inference on the Markov jump process even though its statespace is countably infinite. Numerical experiments show improvements over particle MCMC of between a factor of three and several orders of magnitude.",
keywords = "MCMC, continuous-time Markov chain, coffin state, correlated pseudo-marginal",
author = "Chris Sherlock and Andrew Golightly",
year = "2023",
month = jan,
day = "31",
doi = "10.1080/10618600.2022.2093886",
language = "English",
volume = "32",
pages = "36--48",
journal = "Journal of Computational and Graphical Statistics",
issn = "1061-8600",
publisher = "American Statistical Association",
number = "1",

}

RIS

TY - JOUR

T1 - Exact Bayesian inference for discretely observed Markov Jump Processes using finite rate matrices

AU - Sherlock, Chris

AU - Golightly, Andrew

PY - 2023/1/31

Y1 - 2023/1/31

N2 - We present new methodologies for Bayesian inference on the rate parameters of a discretely observed continuous-time Markov jump process with a countably infinite statespace. The usual method of choice for inference, particle Markov chain Monte Carlo (particle MCMC), struggles when the observation noise is small. We consider the most challenging regime of exact observations and provide two new methodologies for inference in this case: the minimal extended statespace algorithm (MESA) and the nearly minimal extended statespace algorithm (nMESA). By extending the Markov chain Monte Carlo statespace, both MESA and nMESA use the exponentiation of finite rate matrices to perform exact Bayesian inference on the Markov jump process even though its statespace is countably infinite. Numerical experiments show improvements over particle MCMC of between a factor of three and several orders of magnitude.

AB - We present new methodologies for Bayesian inference on the rate parameters of a discretely observed continuous-time Markov jump process with a countably infinite statespace. The usual method of choice for inference, particle Markov chain Monte Carlo (particle MCMC), struggles when the observation noise is small. We consider the most challenging regime of exact observations and provide two new methodologies for inference in this case: the minimal extended statespace algorithm (MESA) and the nearly minimal extended statespace algorithm (nMESA). By extending the Markov chain Monte Carlo statespace, both MESA and nMESA use the exponentiation of finite rate matrices to perform exact Bayesian inference on the Markov jump process even though its statespace is countably infinite. Numerical experiments show improvements over particle MCMC of between a factor of three and several orders of magnitude.

KW - MCMC

KW - continuous-time Markov chain

KW - coffin state

KW - correlated pseudo-marginal

U2 - 10.1080/10618600.2022.2093886

DO - 10.1080/10618600.2022.2093886

M3 - Journal article

VL - 32

SP - 36

EP - 48

JO - Journal of Computational and Graphical Statistics

JF - Journal of Computational and Graphical Statistics

SN - 1061-8600

IS - 1

ER -