Home > Research > Publications & Outputs > Efficient model comparison techniques for model...

Research

Associated organisational units

Electronic data

  • BA1057

    Accepted author manuscript, 479 KB, PDF document

    Available under license: None

Links

Text available via DOI:

Keywords

View graph of relations

Efficient model comparison techniques for models requiring large scale data augmentation

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

Published

Standard

Efficient model comparison techniques for models requiring large scale data augmentation. / Touloupou, Panayiota; Alzahrani, Naif; Neal, Peter John et al.
In: Bayesian Analysis, Vol. 13, No. 2, 03.2018, p. 437-459.

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

Harvard

Touloupou, P, Alzahrani, N, Neal, PJ, Spencer, S & McKinley, T 2018, 'Efficient model comparison techniques for models requiring large scale data augmentation', Bayesian Analysis, vol. 13, no. 2, pp. 437-459. https://doi.org/10.1214/17-BA1057

APA

Touloupou, P., Alzahrani, N., Neal, P. J., Spencer, S., & McKinley, T. (2018). Efficient model comparison techniques for models requiring large scale data augmentation. Bayesian Analysis, 13(2), 437-459. https://doi.org/10.1214/17-BA1057

Vancouver

Touloupou P, Alzahrani N, Neal PJ, Spencer S, McKinley T. Efficient model comparison techniques for models requiring large scale data augmentation. Bayesian Analysis. 2018 Mar;13(2):437-459. Epub 2017 Apr 29. doi: 10.1214/17-BA1057

Author

Touloupou, Panayiota ; Alzahrani, Naif ; Neal, Peter John et al. / Efficient model comparison techniques for models requiring large scale data augmentation. In: Bayesian Analysis. 2018 ; Vol. 13, No. 2. pp. 437-459.

Bibtex

@article{c974da77d899467081ccdfaa37c841cf,
title = "Efficient model comparison techniques for models requiring large scale data augmentation",
abstract = "Selecting between competing statistical models is a challenging problem especially when the competing models are non-nested. In this paper we offer a simple solution by devising an algorithm which combines MCMC and importance sampling to obtain computationally efficient estimates of the marginal likelihood which can then be used to compare the models. The algorithm is successfully applied to a longitudinal epidemic data set, where calculating the marginal likelihood is made more challenging by the presence of large amounts of missing data. In this context, our importance sampling approach is shown to outperform existing methods for computing the marginal likelihood.",
keywords = "Epidemics, marginal likelihood, model evidence, model selection, time series",
author = "Panayiota Touloupou and Naif Alzahrani and Neal, {Peter John} and Simon Spencer and Trevelyan McKinley",
year = "2018",
month = mar,
doi = "10.1214/17-BA1057",
language = "English",
volume = "13",
pages = "437--459",
journal = "Bayesian Analysis",
issn = "1936-0975",
publisher = "Carnegie Mellon University",
number = "2",

}

RIS

TY - JOUR

T1 - Efficient model comparison techniques for models requiring large scale data augmentation

AU - Touloupou, Panayiota

AU - Alzahrani, Naif

AU - Neal, Peter John

AU - Spencer, Simon

AU - McKinley, Trevelyan

PY - 2018/3

Y1 - 2018/3

N2 - Selecting between competing statistical models is a challenging problem especially when the competing models are non-nested. In this paper we offer a simple solution by devising an algorithm which combines MCMC and importance sampling to obtain computationally efficient estimates of the marginal likelihood which can then be used to compare the models. The algorithm is successfully applied to a longitudinal epidemic data set, where calculating the marginal likelihood is made more challenging by the presence of large amounts of missing data. In this context, our importance sampling approach is shown to outperform existing methods for computing the marginal likelihood.

AB - Selecting between competing statistical models is a challenging problem especially when the competing models are non-nested. In this paper we offer a simple solution by devising an algorithm which combines MCMC and importance sampling to obtain computationally efficient estimates of the marginal likelihood which can then be used to compare the models. The algorithm is successfully applied to a longitudinal epidemic data set, where calculating the marginal likelihood is made more challenging by the presence of large amounts of missing data. In this context, our importance sampling approach is shown to outperform existing methods for computing the marginal likelihood.

KW - Epidemics

KW - marginal likelihood

KW - model evidence

KW - model selection

KW - time series

U2 - 10.1214/17-BA1057

DO - 10.1214/17-BA1057

M3 - Journal article

VL - 13

SP - 437

EP - 459

JO - Bayesian Analysis

JF - Bayesian Analysis

SN - 1936-0975

IS - 2

ER -