Nous avons étudié les propriétés asymptotiques de paramètres fonctionnels conditionnels en statistiques non paramétriques, le quantile conditionnel, quand la variable explicative prend ses valeurs dans un espace de dimension infinie. Dans ce

Keywords: machine learning, neural networks, optimization, suboptimal solutions, mathematical games.Abstract. Since the very beginning, there has been a fruitful exchange between machine learning and optimization. While machine learning exploits optimization models and

https://indico.math.cnrs.fr/event/10332/

Given nonempty, closed and convex sets C ⊂ H, Q ⊂ K in Hilbert spacesH, K and a bounded linear operator A : H → K, the split feasibility problem (SFP) is

"A coupling approach to the turnpike problem in stochastic control"The turnpike principle is a general phenomenon in dynamic control postulating that optimal curves and control tend to spend most of their time

In this presentation, we address optimal control problems involving loss control regions. In this context, the state space is partitioned into disjoint sets referred to as regions, which are classified into two

Finite mixtures of probability distributions provide a powerful and flexible mechanism for modeling real-life data that cannot be easily approximated by elementary distributions. Among other applications, mixture models found their use in

Cet exposé se concentre sur l'analyse des jeux différentiels à sommes nulles impliquant deux joueurs soumis à des dynamiques continues indépendantes et des contraintes d'états distinctes. Les interactions entre ces joueurs sont

This presentation focuses on the estimation of the variance function in regression and its applications in regression with reject option and prediction intervals.First, we are interested in estimating the variance function through

The joint parametric estimation of the drift coefficient, the scale coefficient and the jump activity in stochastic differential equations driven by a symmetric stable Lévy process is considered, based on high-frequency observations.Firstly,

The objective of this talk is to present a stochastic maximum principle for a control problem with a criterion implicitly defined by a quadratic BSDE. Quadratic growth BSDEs are well studied objects and

Résumé : L'inférence des quantiles extrêmes est importante pour l'évaluation des risques, par exemple dans les applications hydrologiques, où les valeurs rapportées sont utilisées pour élaborer des plans de prévention contre les

Imagine you want to solve a maze. Easy peasy, Dijkstra's shortest path algorithm gets the job done.Now imagine the maze is not given to you as a nice clean graph: instead, you

Many problems in imaging and data science require to reconstruct, from partial observations, highly concentrated signals, such as pointwise sources or contour lines. This work introduces a novel algorithm for recovering measures

Hawkes process was introduced by Hawkes in 1971 and are widely used in many applications (earthquakes, neurons, social network, finance, etc.). This jump process has an intensity which depends on the past.

Le sujet de cette présentation est un problème inverse, qui consiste à reconstruire une image à partir de mesures linéaires bruitées. Je traiterai plus spécifiquement d'une méthode de reconstruction variationnelle, qui utilise

This work addresses second-order stochastic optimization for estimating the minimizer of a convex function written as an expectation. A direct recursive estimation technique for the inverse Hessian matrix using a Robbins-Monro procedure

Les systèmes optiques sont constamment améliorés, afin d'augmenter la résolution, de réduire le temps d'acquisition ou simplement de permettre l'observation d'échantillons qui ne peuvent être imagés avec les systèmes actuels. Les gains

We address a large-scale and non-convex optimization problem, involving an aggregative term. This term can be interpreted as the sum of the contributions of N agents to some common good, with N

Dans cet exposé, je vais aborder l'inférence statistique en discutant de méthodes d'estimations comme la méthode des fonctions caractéristiques, l'algorithme EM, l’approche Bayésienne, statistiques des extrêmes dans le contexte des lois stables. Ensuite, je présenterai une

Nous commencerons par examiner le problème du temps de sortie, en décrivant sa formulation et les résultats établis dans le contexte de la diffusion d'Itô, connue sous le nom de théorie de

In this talk, we present the different issues to design of a smart electrostimu- lator for muscle rehabilitation or reinforcement, in order to control the muscular force or to regulate to a

Je commencerai par présenter le contexte industrielle qui nous à amener à développer une méthode de lissage de spectre infrarouge. Puis nous introduirons les bases théoriques nécessaires à l'analyse topologique de données.

Le modèle d'agrégation limitée par diffusion interne (IDLA) est un modèle de croissance aléatoire sur Z^d. Ce dernier consiste en une famille d'agrégats aléatoires qui sont définis de façon récursive comme suit:

L'estimateur SLOPE, acronyme signifiant « Sorted L One Penalized Estimation », est défini comme une solution d'un problème d'optimisation convexe où le terme de pénalité est la norme L1 ordonnée. Cette norme,

We propose a novel notion of viscosity solutions to study first order Hamilton-Jacobi equa- tions in a certain class of metric spaces called geodesic spaces of one curvature bound in the sense

A classical problem in ergodic control consists of studying the limit behaviour of the optimal value of a discounted cost functional with infinite horizon as the discount fact tends to zero. In

Résumé : In this presentation, we explore three non-convex optimization problems related to the training of shallow neural networks (NNs) for exact and approximate representation, as well as regression tasks. From a

In this talk, we consider the class of optimal control problems, where the control system is given by a controlled second order differential equation. Such control systems naturally appear in mechanics in

Abstract: Ranking random observations has become essential to many data analysis problems, ranging from recommendation systems, computational biology, to information retrieval for instance. The information acquisition processes often involve various and poorly

An important challenge in DNA replication analysis is to recover the so-called replication timing profile, that contains important information about the DNA replication dynamics. We show that using recent replication data, this

Coupling Brownian motions in a useful tool to obtain results in Probability, Geometry and Analysis. In particular, by studying the first meeting time of Brownian motions, we can obtain gradient inequalities for

Due to the fast growth of data that are measured on a continuous scale, functional data analysis has undergone manydevelopments in recent years. Regression models with a functional response involving functional covariates,

Abstract: In this talk, we will adress the following question: how can we enforce both parsimony and structure in a regularized regression model ? Specifically, we investigate strategies that promote not only

9h HERVE CARDOT Titre : Statistical modeling of categorical trajectories with multivariate functional data approaches9h45 BERNARD BONNARD Titre : Problème Géométrique de Zermelo et Micromagnétisme10h15 KANG LIU Titre : Universal approximation of dynamical systems by

It is well known that Earth's climatic characteristics have undergone significant variations throughout the planet's history. Even today, we are witnessing increasingly rapid and intense climate changes, likely driven by human intervention

We introduce viscosity solutions to some monotone nonlinear PDEs, and the particular case of control problems. This class of equations can be extended to some metric spaces, provided one has well-posedness of

Le problème d'optimisation que nous allons étudier est motivé par la modélisation de la dynamique des habitants dans une ville où les agents déménagent d’un endroit à un autre. La stratégie d'un

Résumé : We study the volume of rigid loop-O(n) quadrangulations with a boundary of length 2p in the critical non-generic regime. We prove that, as the half-perimeter p goes to infinity, the volume scales in distribution to

Résumé : Dans cet exposé, je présenterai un nouveau modèle d'arbres aléatoires qui est une généralisation naturelle des arbres de Bienaymé-Galton-Watson (BGW), et où les morts des individus deviennent spatialement corrélées, via