Skip to main content

A loss-based prior for variable selection in linear regression methods

Villa, Cristiano, Lee, Jeong Eun (2019) A loss-based prior for variable selection in linear regression methods. Bayesian Analysis, . ISSN 1936-0975. E-ISSN 1931-6690. (doi:10.1214/19-BA1162)

PDF - Publisher pdf

Creative Commons Licence
This work is licensed under a Creative Commons Attribution 4.0 International License.
Download (521kB) Preview
[img]
Preview
PDF - Author's Accepted Manuscript
Restricted to Repository staff only
Contact us about this Publication Download (372kB)
[img]
PDF - Author's Accepted Manuscript
Download (461kB) Preview
[img]
Preview
Official URL
http://dx.doi.org/10.1214/19-BA1162

Abstract

In this work we propose a novel model prior for variable selection in linear regression. The idea is to determine the prior mass by considering the \emph{worth} of each of the regression models, given the number of possible covariates under consideration. The worth of a model consists of the information loss and the loss due to model complexity. While the information loss is determined objectively, the loss expression due to model complexity is flexible and, the penalty on model size can be even customized to include some prior knowledge. Some versions of the loss-based prior are proposed and compared empirically. Through simulation studies and real data analyses, we compare the proposed prior to the Scott and Berger prior, for noninformative scenarios, and with the Beta-Binomial prior, for informative scenarios.

Item Type: Article
DOI/Identification number: 10.1214/19-BA1162
Subjects: Q Science > QA Mathematics (inc Computing science) > QA276 Mathematical statistics
Divisions: Faculties > Sciences > School of Mathematics Statistics and Actuarial Science > Statistics
Depositing User: Cristiano Villa
Date Deposited: 01 May 2019 05:51 UTC
Last Modified: 14 Aug 2019 08:40 UTC
Resource URI: https://kar.kent.ac.uk/id/eprint/73709 (The current URI for this page, for reference purposes)
Villa, Cristiano: https://orcid.org/0000-0002-2670-2954
  • Depositors only (login required):

Downloads

Downloads per month over past year