93
28

A Bernstein-type inequality for suprema of random processes with applications to model selection in non-Gaussian regression

Abstract

Let \paXttT\pa{X_{t}}_{t\in T} be a family of real-valued centered random variables indexed by a countable set TT. In the first part of this paper, we establish exponential bounds for the deviation probabilities of the supremum Z=suptTXtZ=\sup_{t\in T}X_{t} by using the generic chaining device introduced in Talagrand (2005). Compared to concentration-type inequalities, these bounds offer the advantage to hold under weaker conditions on the family \paXttT\pa{X_{t}}_{t\in T}. The second part of the paper is oriented towards statistics. We consider the regression setting Y=f+\epsY=f+\eps where ff is an unknown vector of Rn\R^{n} and \eps\eps is a random vector the components of which are independent, centered and admit finite Laplace transforms in a neighborhood of 0. Our aim is to estimate ff from the observation of YY by mean of a model selection approach among a collection of linear subspaces of Rn\R^{n}. The selection procedure we propose is based on the minimization of a penalized criterion the penalty of which is calibrated by using the deviation bounds established in the first part of this paper. More precisely, we study suprema of random variables of the form Xt=i=1nti\epsiX_{t}=\sum_{i=1}^{n}t_{i}\eps_{i} when tt varies among the unit ball of a linear subspace of Rn\R^{n}. We finally show that our estimator satisfies some oracle-type inequality under suitable assumptions on the metric structures of the linear spaces of the collection.

View on arXiv
Comments on this paper