Shrinkage priors for Bayesian prediction

Abstract

We investigate shrinkage priors for constructing Bayesian predictive distributions. It is shown that there exist shrinkage predictive distributions asymptotically dominating Bayesian predictive distributions based on the Jeffreys prior or other vague priors if the model manifold satisfies some differential geometric conditions. Kullback--Leibler divergence from the true distribution to a predictive distribution is adopted as a loss function. Conformal transformations of model manifolds corresponding to vague priors are introduced. We show several examples where shrinkage predictive distributions dominate Bayesian predictive distributions based on vague priors.

Published at http://dx.doi.org/10.1214/009053606000000010 in the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (3)

• Improved minimax predictive densities under Kullback--Leibler loss
• Shrinkage priors for Bayesian prediction
• Simultaneous prediction of independent Poisson observables

Cited by in corpus (16)

• Parametric bootstrap approximation to the distribution of EBLUP and related prediction intervals in linear mixed models
• Shrinkage priors for Bayesian prediction
• Simultaneous prediction of independent Poisson observables
• From Minimax Shrinkage Estimation to Minimax Shrinkage Prediction
• Asymptotic Properties of Bayesian Predictive Densities When the Distributions of Data and Target Variables are Different
• Empirical geodesic graphs and CAT(k) metrics for data analysis
• Kählerian information geometry for signal processing
• Asymptotic Expansion of the Risk Difference of the Bayesian Spectral Density in the ARMA model
• Application of Kähler manifold to signal processing and Bayesian inference
• Geometric shrinkage priors for Kählerian signal filters
• Kähler information manifolds of signal processing filters in weighted Hardy spaces
• Nearly minimax empirical Bayesian prediction of independent Poisson observables
• Bayesian shrinkage prediction for the regression problem
• Estimation under matrix quadratic loss and matrix superharmonicity
• Shrinkage priors on complex-valued circular-symmetric autoregressive processes
• Inadmissibility of the best equivariant predictive density in the unknown variance case