Home > Research > Publications & Outputs > Markov chain Monte Carlo for exact inference fo...

Research

Associated organisational unit

Links

Text available via DOI:

Keywords

View graph of relations

Markov chain Monte Carlo for exact inference for diffusions

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

Published

Standard

Markov chain Monte Carlo for exact inference for diffusions. / Sermaidis, Giorgos; Papaspiliopoulos, Omiros; Roberts, Gareth et al.
In: Scandinavian Journal of Statistics, Vol. 40, No. 2, 06.2013, p. 294-321.

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

Harvard

Sermaidis, G, Papaspiliopoulos, O, Roberts, G, Beskos, A & Fearnhead, P 2013, 'Markov chain Monte Carlo for exact inference for diffusions', Scandinavian Journal of Statistics, vol. 40, no. 2, pp. 294-321. https://doi.org/10.1111/j.1467-9469.2012.00812.x

APA

Sermaidis, G., Papaspiliopoulos, O., Roberts, G., Beskos, A., & Fearnhead, P. (2013). Markov chain Monte Carlo for exact inference for diffusions. Scandinavian Journal of Statistics, 40(2), 294-321. https://doi.org/10.1111/j.1467-9469.2012.00812.x

Vancouver

Sermaidis G, Papaspiliopoulos O, Roberts G, Beskos A, Fearnhead P. Markov chain Monte Carlo for exact inference for diffusions. Scandinavian Journal of Statistics. 2013 Jun;40(2):294-321. doi: 10.1111/j.1467-9469.2012.00812.x

Author

Sermaidis, Giorgos ; Papaspiliopoulos, Omiros ; Roberts, Gareth et al. / Markov chain Monte Carlo for exact inference for diffusions. In: Scandinavian Journal of Statistics. 2013 ; Vol. 40, No. 2. pp. 294-321.

Bibtex

@article{54dba676c58445e19ef81720bf7a4d0c,
title = "Markov chain Monte Carlo for exact inference for diffusions",
abstract = "We develop exact Markov chain Monte Carlo methods for discretely sampled, directly and indirectly observed diffusions. The qualification {\textquoteleft}exact{\textquoteright} refers to the fact that the invariant and limiting distribution of the Markov chains is the posterior distribution of the parameters free of any discretization error. The class of processes to which our methods directly apply are those which can be simulated using the most general to date exact simulation algorithm. The article introduces various methods to boost the performance of the basic scheme, including reparametrizations and auxiliary Poisson sampling. We contrast both theoretically and empirically how this new approach compares to irreducible high frequency imputation, which is the state-of-the-art alternative for the class of processes we consider, and we uncover intriguing connections. All methods discussed in the article are tested on typical examples.",
keywords = "exact inference, xact simulation, transition density, ;Markov chain Monte Carlo, differential equation, stochastic ",
author = "Giorgos Sermaidis and Omiros Papaspiliopoulos and Gareth Roberts and Alexandros Beskos and Paul Fearnhead",
year = "2013",
month = jun,
doi = "10.1111/j.1467-9469.2012.00812.x",
language = "English",
volume = "40",
pages = "294--321",
journal = "Scandinavian Journal of Statistics",
issn = "0303-6898",
publisher = "Blackwell-Wiley",
number = "2",

}

RIS

TY - JOUR

T1 - Markov chain Monte Carlo for exact inference for diffusions

AU - Sermaidis, Giorgos

AU - Papaspiliopoulos, Omiros

AU - Roberts, Gareth

AU - Beskos, Alexandros

AU - Fearnhead, Paul

PY - 2013/6

Y1 - 2013/6

N2 - We develop exact Markov chain Monte Carlo methods for discretely sampled, directly and indirectly observed diffusions. The qualification ‘exact’ refers to the fact that the invariant and limiting distribution of the Markov chains is the posterior distribution of the parameters free of any discretization error. The class of processes to which our methods directly apply are those which can be simulated using the most general to date exact simulation algorithm. The article introduces various methods to boost the performance of the basic scheme, including reparametrizations and auxiliary Poisson sampling. We contrast both theoretically and empirically how this new approach compares to irreducible high frequency imputation, which is the state-of-the-art alternative for the class of processes we consider, and we uncover intriguing connections. All methods discussed in the article are tested on typical examples.

AB - We develop exact Markov chain Monte Carlo methods for discretely sampled, directly and indirectly observed diffusions. The qualification ‘exact’ refers to the fact that the invariant and limiting distribution of the Markov chains is the posterior distribution of the parameters free of any discretization error. The class of processes to which our methods directly apply are those which can be simulated using the most general to date exact simulation algorithm. The article introduces various methods to boost the performance of the basic scheme, including reparametrizations and auxiliary Poisson sampling. We contrast both theoretically and empirically how this new approach compares to irreducible high frequency imputation, which is the state-of-the-art alternative for the class of processes we consider, and we uncover intriguing connections. All methods discussed in the article are tested on typical examples.

KW - exact inference

KW - xact simulation

KW - transition density

KW - ;Markov chain Monte Carlo

KW - differential equation

KW - stochastic

U2 - 10.1111/j.1467-9469.2012.00812.x

DO - 10.1111/j.1467-9469.2012.00812.x

M3 - Journal article

VL - 40

SP - 294

EP - 321

JO - Scandinavian Journal of Statistics

JF - Scandinavian Journal of Statistics

SN - 0303-6898

IS - 2

ER -