Polynomial Log-Marginals and Tweedie's Formula : When Is Bayes Possible?

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Datta, Jyotishka, Polson, Nicholas G.
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917224327938048
author Datta, Jyotishka
Polson, Nicholas G.
author_facet Datta, Jyotishka
Polson, Nicholas G.
contents Motivated by Tweedie's formula for the Compound Decision problem, we examine the theoretical foundations of empirical Bayes estimators that directly model the marginal density $m(y)$. Our main result shows that polynomial log-marginals of degree $k \ge 3 $ cannot arise from any valid prior distribution in exponential family models, while quadratic forms correspond exactly to Gaussian priors. This provides theoretical justification for why certain empirical Bayes decision rules, while practically useful, do not correspond to any formal Bayes procedures. We also strengthen the diagnostic by showing that a marginal is a Gaussian convolution only if it extends to a bounded solution of the heat equation in a neighborhood of the smoothing parameter, beyond the convexity of $c(y)=\tfrac12 y^2+\log m(y)$.
format Preprint
id arxiv_https___arxiv_org_abs_2509_05823
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Polynomial Log-Marginals and Tweedie's Formula : When Is Bayes Possible?
Datta, Jyotishka
Polson, Nicholas G.
Statistics Theory
Econometrics
Methodology
62C12, 60E10, 62C25
Motivated by Tweedie's formula for the Compound Decision problem, we examine the theoretical foundations of empirical Bayes estimators that directly model the marginal density $m(y)$. Our main result shows that polynomial log-marginals of degree $k \ge 3 $ cannot arise from any valid prior distribution in exponential family models, while quadratic forms correspond exactly to Gaussian priors. This provides theoretical justification for why certain empirical Bayes decision rules, while practically useful, do not correspond to any formal Bayes procedures. We also strengthen the diagnostic by showing that a marginal is a Gaussian convolution only if it extends to a bounded solution of the heat equation in a neighborhood of the smoothing parameter, beyond the convexity of $c(y)=\tfrac12 y^2+\log m(y)$.
title Polynomial Log-Marginals and Tweedie's Formula : When Is Bayes Possible?
topic Statistics Theory
Econometrics
Methodology
62C12, 60E10, 62C25
url https://arxiv.org/abs/2509.05823