Journal Article (Journal Article)

We propose a generalized double Pareto prior for Bayesian shrinkage estimation and inferences in linear models. The prior can be obtained via a scale mixture of Laplace or normal distributions, forming a bridge between the Laplace and Normal-Jeffreys' priors. While it has a spike at zero like the Laplace density, it also has a Student's t -like tail behavior. Bayesian computation is straightforward via a simple Gibbs sampling algorithm. We investigate the properties of the maximum a posteriori estimator, as sparse estimation plays an important role in many problems, reveal connections with some well-established regularization procedures, and show some asymptotic results. The performance of the prior is tested through simulations and an application.

Full Text

Duke Authors

Cited Authors

  • Armagan, A; Dunson, DB; Lee, J

Published Date

  • January 2013

Published In

Volume / Issue

  • 23 / 1

Start / End Page

  • 119 - 143

PubMed ID

  • 24478567

Pubmed Central ID

  • PMC3903426

Electronic International Standard Serial Number (EISSN)

  • 1996-8507

International Standard Serial Number (ISSN)

  • 1017-0405


  • eng