Shrinkage priors for Bayesian prediction

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.

[1]  Seizô Itô.  Martin boundary for linear elliptic differential operators of second order in a manifold , 1964 .

[2]  J. Aitchison,et al.  Statistical Prediction Analysis , 1975 .

[3]  Malay Ghosh,et al.  Bayesian prediction with approximate frequentist validity , 2000 .

[4]  L. Brown.  A Heuristic Method for Determining Admissibility of Estimators--With Applications , 1979 .

[5]  H. Urakawa.  Geometry of Laplace–Beltrami Operator on a Complete Riemannian Manifold , 1993 .

[6]  F. Götze.  Differential-geometrical methods in statistics. Lecture notes in statistics - A. Shun-ichi. , 1987 .

[7]  Fumiyasu Komaki.  Bayesian Predictive Distribution with Right Invariant Priors , 2002 .

[8]  E. Davies,et al.  Heat kernels and spectral theory , 1989 .

[9]  L. Brown.  Admissible Estimators, Recurrent Diffusions, and Insoluble Boundary Value Problems , 1971 .

[10]  Fumiyasu Komaki,et al.  Simultaneous prediction of independent Poisson observables , 2004 .

[11]  Second order availability and positive solutions of the Schrödinger equation , 1983 .

[12]  B. Levit.  Second-Order Asymptotic Optimality and Positive Solutions of Schrödinger’s Equation , 1986 .

[13]  F. Komaki.  On asymptotic properties of predictive distributions , 1996 .

[14]  Shun-ichi Amari,et al.  Differential-geometrical methods in statistics , 1985 .

[15]  Seizô Itô.  On existence of Green function and positive superharmonic functions for linear elliptic operators of second order , 1964 .

[16]  Jayanta K. Ghosh,et al.  Higher Order Asymptotics , 1994 .

[17]  Paolo Vidoni,et al.  A simple predictive density based on the p*-formula , 1995 .

[18]  Shun-ichi Amari,et al.  Methods of information geometry , 2000 .

[19]  C. Stein.  Estimation of the Mean of a Multivariate Normal Distribution , 1981 .

[20]  Feng Liang,et al.  Improved minimax predictive densities under Kullback-Leibler loss , 2006 .

[21]  Fumiyasu Komaki,et al.  A shrinkage predictive distribution for multivariate Normal observables , 2001 .

[22]  I. Holopainen.  Riemannian Geometry , 1927, Nature.

[23]  K. Aomoto.  L'analyse harmonique sur les espaces riemanniens, a courbure riemannienne negative I , 1966 .

[24]  J. Hartigan.  The maximum likelihood prior , 1998 .

[25]  Seymour Geisser,et al.  8. Predictive Inference: An Introduction , 1995 .

[26]  J. Zidek.  A representation of Bayes invariant procedures in terms of Haar measure , 1969 .

[27]  W. H. Williams,et al.  Probability Theory and Mathematical Statistics , 1964 .

[28]  William E. Strawderman,et al.  Generalizations of James-Stein Estimators Under Spherical Symmetry , 1991 .