Originally, oracle inequalities were developed as particularly efficient tools in mathematical statistics for deriving adaptation to an unknown parameter. Initially dedicated to the demonstration of statistical properties in the study of certain estimators, they can be extrapolated to the broader framework of aggregation in which they constitute the core of the definition of an optimal rate of aggregation. As such, they become not only mathematical tools but also precise finite sample results. In this thesis, several fields of application of oracle inequalities are presented. They are used first in a general aggregation framework and then in particular statistical models such as nonparametric density estimation, regression and classificatio...
Abstract. This paper studies oracle properties of `1-penalized estima-tors of a probability density....
The paper presents recent developments of the theory of estimator selection. We introduce,...
This manuscript focuses on two functional estimation problems. A non asymptotic guarantee of the pro...
Originally, oracle inequalities were developed as particularly efficient tools in mathematical stati...
We treat two subjects. The first subject is about statistical learning in high-dimension, that is wh...
Ce manuscrit se concentre sur deux problèmes d'estimation de fonction. Pour chacun, une garantie non...
Since univariate extreme value theory is essentially based on asymptotic theorems, the elaboration o...
Deux sujet sont traités dans cette thèse: l'agrégation d'estimateurs et la régression sous contraint...
À partir des observations Z(n) = {(Xi, Yi), i = 1, ..., n} satisfaisant Yi = f(Xi) + ζi, nous voulon...
Dans cette thèse nous traitons deux sujets. Le premier sujet concerne l'apprentissage statistique en...
International audienceWe consider the problem of combining a (possibly uncountably infinite) set of ...
In one-dimensional density estimation on i.i.d. observations we suggest an adaptive cross-validation...
The purpose of these lecture notes is to provide an introduction to the general theory of empirical ...
International audienceAggregating estimators using exponential weights depending on their risk perfo...
International audienceWe study the problem of nonparametric estimation under L p-loss, p ∈ [1, ∞), i...
Abstract. This paper studies oracle properties of `1-penalized estima-tors of a probability density....
The paper presents recent developments of the theory of estimator selection. We introduce,...
This manuscript focuses on two functional estimation problems. A non asymptotic guarantee of the pro...
Originally, oracle inequalities were developed as particularly efficient tools in mathematical stati...
We treat two subjects. The first subject is about statistical learning in high-dimension, that is wh...
Ce manuscrit se concentre sur deux problèmes d'estimation de fonction. Pour chacun, une garantie non...
Since univariate extreme value theory is essentially based on asymptotic theorems, the elaboration o...
Deux sujet sont traités dans cette thèse: l'agrégation d'estimateurs et la régression sous contraint...
À partir des observations Z(n) = {(Xi, Yi), i = 1, ..., n} satisfaisant Yi = f(Xi) + ζi, nous voulon...
Dans cette thèse nous traitons deux sujets. Le premier sujet concerne l'apprentissage statistique en...
International audienceWe consider the problem of combining a (possibly uncountably infinite) set of ...
In one-dimensional density estimation on i.i.d. observations we suggest an adaptive cross-validation...
The purpose of these lecture notes is to provide an introduction to the general theory of empirical ...
International audienceAggregating estimators using exponential weights depending on their risk perfo...
International audienceWe study the problem of nonparametric estimation under L p-loss, p ∈ [1, ∞), i...
Abstract. This paper studies oracle properties of `1-penalized estima-tors of a probability density....
The paper presents recent developments of the theory of estimator selection. We introduce,...
This manuscript focuses on two functional estimation problems. A non asymptotic guarantee of the pro...