Guardado en:
Detalles Bibliográficos
Autores principales: Liu, Chang, Martin, Ryan, Shen, Weining
Formato: Preprint
Publicado: 2017
Materias:
Acceso en línea:https://arxiv.org/abs/1712.03848
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909716412628992
author Liu, Chang
Martin, Ryan
Shen, Weining
author_facet Liu, Chang
Martin, Ryan
Shen, Weining
contents Inference on high-dimensional parameters in structured linear models is an important statistical problem. This paper focuses on the case of a piecewise polynomial Gaussian sequence model, and we develop a new empirical Bayes solution that enjoys adaptive minimax posterior concentration rates and improved structure learning properties compared to existing methods. Moreover, thanks to the conjugate form of the empirical prior, posterior computations are fast and easy. Numerical examples also highlight the method's strong finite-sample performance compared to existing methods across a range of different scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_1712_03848
institution arXiv
publishDate 2017
record_format arxiv
spellingShingle Empirical priors and posterior concentration in a piecewise polynomial sequence model
Liu, Chang
Martin, Ryan
Shen, Weining
Statistics Theory
Inference on high-dimensional parameters in structured linear models is an important statistical problem. This paper focuses on the case of a piecewise polynomial Gaussian sequence model, and we develop a new empirical Bayes solution that enjoys adaptive minimax posterior concentration rates and improved structure learning properties compared to existing methods. Moreover, thanks to the conjugate form of the empirical prior, posterior computations are fast and easy. Numerical examples also highlight the method's strong finite-sample performance compared to existing methods across a range of different scenarios.
title Empirical priors and posterior concentration in a piecewise polynomial sequence model
topic Statistics Theory
url https://arxiv.org/abs/1712.03848