Improved minimax predictive densities under Kullback-Leibler loss

Let X|μ ∼ N p (μ, u x I) and Y|μ ∼ N p (μ, v y I) be independentp-dimensional multivariate normal vectors with common unknown mean μ. Based on only observing X = x, we consider the problem of obtaining a predictive density p(y|x) for Y that is close to p(y|μ) as measured by expected Kullback-Leibler loss. A natural procedure for this problem is the (formal) Bayes predictive density p U (y|x) under the uniform prior π U (μ) ≡ 1, which is best invariant and minimax. We show that any Bayes predictive density will be minimax if it is obtained by a prior yielding a marginal that is superharmonic or whose square root is superharmonic. This yields wide classes of minimax procedures that dominate p U (y|x), including Bayes predictive densities under superharmonic priors. Fundamental similarities and differences with the parallel theory of estimating a multivariate normal mean under quadratic loss are described.

[1]  Fumiyasu Komaki,et al.  Simultaneous prediction of independent Poisson observables , 2004, math/0410094.

[2]  E. L. Lehmann,et al.  Theory of point estimation , 1950 .

[3]  Mihaela Aslan,et al.  Asymptotically minimax Bayes predictive densities , 2006, 0708.0177.

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

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

[6]  Malay Ghosh,et al.  Nonsubjective priors via predictive relative entropy regret , 2006 .

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

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

[9]  E. George.  Combining Minimax Shrinkage Estimators , 1986 .

[10]  G. D. Murray,et al.  NOTE ON ESTIMATION OF PROBABILITY DENSITY FUNCTIONS , 1977 .

[11]  William E. Strawderman,et al.  The Heat Equation and Stein's Identity: Connections, Applications , 2006 .

[12]  E. George.  A formal bayes multiple shrinkage estimator , 1986 .

[13]  Ian R. Harris.  Predictive fit for natural exponential families , 1989 .

[14]  Vee Ming Ng,et al.  On the estimation of parametric density functions , 1980 .

[15]  J. Aitchison.  Goodness of prediction fit , 1975 .

[16]  J. Steele.  Stochastic Calculus and Financial Applications , 2000 .

[17]  E. George.  Minimax Multiple Shrinkage Estimation , 1986 .

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

[19]  W. Strawderman.  Proper Bayes Minimax Estimators of the Multivariate Normal Mean , 1971 .

[20]  Feng Liang,et al.  Exact minimax strategies for predictive density estimation, data compression, and model selection , 2002, IEEE Transactions on Information Theory.

[21]  J. Berger.  Statistical Decision Theory and Bayesian Analysis , 1988 .

[22]  M. Wells,et al.  On the construction of Bayes minimax estimators , 1998 .