On the Choice of Prior in Bayesian Model Averaging

J.H.J. Einmahl, J.R. Magnus, K. Kumar

Research output: Working paperDiscussion paperOther research output

285 Downloads (Pure)

Abstract

Bayesian model averaging attempts to combine parameter estimation and model uncertainty in one coherent framework. The choice of prior is then critical. Within an explicit framework of ignorance we define a ‘suitable’ prior as one which leads to a continuous and suitable analog to the pretest estimator. The normal prior, used in standard Bayesian model averaging, is shown to be unsuitable. The Laplace (or lasso) prior is almost suitable. A suitable prior (the Subbotin prior) is proposed and its properties are investigated.
Original languageEnglish
Place of PublicationTILBURG
PublisherEconometrics
Number of pages29
Volume2011-003
Publication statusPublished - 2011

Publication series

NameCentER Discussion Paper
Volume2011-003

Keywords

  • Model averaging
  • Bayesian analysis
  • Subbotin prior

Fingerprint Dive into the research topics of 'On the Choice of Prior in Bayesian Model Averaging'. Together they form a unique fingerprint.

  • Cite this

    Einmahl, J. H. J., Magnus, J. R., & Kumar, K. (2011). On the Choice of Prior in Bayesian Model Averaging. (CentER Discussion Paper; Vol. 2011-003). Econometrics.