Simultaneous parameter estimation and variable selection via the logit-normal continuous analogue of the spike-and-slab prior

Research output: Contribution to journalArticle

Standard

Harvard

APA

Vancouver

Author

Bibtex

@article{f9d37a7782394e6b8de698e95583c773,
title = "Simultaneous parameter estimation and variable selection via the logit-normal continuous analogue of the spike-and-slab prior",
abstract = "We introduce a Bayesian prior distribution, the logit-normal continuous analogue of the spike-and-slab, which enables flexible parameter estimation and variable/model selection in a variety of settings. We demonstrate its use and efficacy in three case studies—a simulation study and two studies on real biological data from the fields of metabolomics and genomics. The prior allows the use of classical statistical models, which are easily interpretable and well known to applied scientists, but performs comparably to common machine learning methods in terms of generalizability to previously unseen data.",
keywords = "Bayesian, Shrinkage, Spike-and-slab, Variable selection",
author = "W. Thomson and S. Jabbari and Taylor, {A. E.} and W. Arlt and David Smith",
year = "2019",
month = jan,
day = "31",
doi = "10.1098/rsif.2018.0572",
language = "English",
volume = "16",
pages = "20180572",
journal = "Journal of The Royal Society Interface",
issn = "1742-5689",
publisher = "The Royal Society",
number = "150",

}

RIS

TY - JOUR

T1 - Simultaneous parameter estimation and variable selection via the logit-normal continuous analogue of the spike-and-slab prior

AU - Thomson, W.

AU - Jabbari, S.

AU - Taylor, A. E.

AU - Arlt, W.

AU - Smith, David

PY - 2019/1/31

Y1 - 2019/1/31

N2 - We introduce a Bayesian prior distribution, the logit-normal continuous analogue of the spike-and-slab, which enables flexible parameter estimation and variable/model selection in a variety of settings. We demonstrate its use and efficacy in three case studies—a simulation study and two studies on real biological data from the fields of metabolomics and genomics. The prior allows the use of classical statistical models, which are easily interpretable and well known to applied scientists, but performs comparably to common machine learning methods in terms of generalizability to previously unseen data.

AB - We introduce a Bayesian prior distribution, the logit-normal continuous analogue of the spike-and-slab, which enables flexible parameter estimation and variable/model selection in a variety of settings. We demonstrate its use and efficacy in three case studies—a simulation study and two studies on real biological data from the fields of metabolomics and genomics. The prior allows the use of classical statistical models, which are easily interpretable and well known to applied scientists, but performs comparably to common machine learning methods in terms of generalizability to previously unseen data.

KW - Bayesian

KW - Shrinkage

KW - Spike-and-slab

KW - Variable selection

UR - http://www.scopus.com/inward/record.url?scp=85061315052&partnerID=8YFLogxK

U2 - 10.1098/rsif.2018.0572

DO - 10.1098/rsif.2018.0572

M3 - Article

C2 - 30958174

VL - 16

SP - 20180572

JO - Journal of The Royal Society Interface

JF - Journal of The Royal Society Interface

SN - 1742-5689

IS - 150

M1 - 20180572

ER -