# High-dimensional Gaussian model selection on a Gaussian design

2 SELECT - Model selection in statistical learning
LMO - Laboratoire de Mathématiques d'Orsay, Inria Saclay - Ile de France
Abstract : We consider the problem of estimating the conditional mean of a real Gaussian variable $\nolinebreak Y=\sum_{i=1}^p\nolinebreak\theta_iX_i+\nolinebreak \epsilon$ where the vector of the covariates $(X_i)_{1\leq i\leq p}$ follows a joint Gaussian distribution. This issue often occurs when one aims at estimating the graph or the distribution of a Gaussian graphical model. We introduce a general model selection procedure which is based on the minimization of a penalized least-squares type criterion. It handles a variety of problems such as ordered and complete variable selection, allows to incorporate some prior knowledge on the model and applies when the number of covariates $p$ is larger than the number of observations $n$. Moreover, it is shown to achieve a non-asymptotic oracle inequality independently of the correlation structure of the covariates. We also exhibit various minimax rates of estimation in the considered framework and hence derive adaptiveness properties of our procedure.
Keywords :
Document type :
Preprints, Working Papers, ...
Domain :
Complete list of metadata

https://hal.inria.fr/inria-00311412
Contributor : Nicolas Verzelen <>
Submitted on : Tuesday, April 28, 2009 - 10:48:34 AM
Last modification on : Wednesday, October 14, 2020 - 4:21:35 AM
Long-term archiving on: : Wednesday, September 22, 2010 - 12:55:24 PM

### Files

RR-oracle.pdf
Files produced by the author(s)

### Identifiers

• HAL Id : inria-00311412, version 2
• ARXIV : 0808.2152

### Citation

Nicolas Verzelen. High-dimensional Gaussian model selection on a Gaussian design. 2008. ⟨inria-00311412v2⟩

Record views