Séminaire Analyse-Probabilités

---

GIRALT Mar (Observatoire de Paris )

Le 17/03/2026
De 10:30 à 11:30
Titre : On the Arnold diffusion mechanism in Medium Earth Orbit

Résumé : Arnold diffusion is a phenomenon in Hamiltonian dynamics in which small perturbations of nearly integrable systems can induce slow but unbounded changes in the action variables. First described by V.I. Arnold (1964), it allows certain trajectories to drift across phase space, despite the perturbations being arbitrarily small. In this talk, we will introduce the Arnold diffusion mechanism and illustrate an application to orbital mechanics. Focusing on the Medium Earth Orbit (MEO) region, home to navigation satellites such as GPS and Galileo, we show how natural perturbations (in our case, third-body effects) can be exploited to guide satellites toward atmospheric reentry at the end of their operational life. Using Galileo as a case study, we analyze a hierarchy of Hamiltonian models incorporating the Earth’s oblateness and the Moon’s gravitational attraction. We demonstrate how Arnold diffusion can trigger eccentricity growth along certain trajectories, lowering the satellite into the atmospheric drag domain. This is a joint work with E.M Alessi, I. Baldomá and M. Guardia.

Salle : A711

Groupe de travail Probabilites

---

VIGOLO Lorenzo ()

Le 18/03/2026
De 14:00 à 15:15
Titre : Introduction to Local Multiple SLEs

Résumé : SLE processes where introduced by Schramm in the 2000 who conjecture and later proved with Lawler and Werner in 2004 that they are the only possible processes that can describe the scaling limits of a random interface in critical 2D statistical mechanics models. He discovered that they are classified by a single parameter $kappa > 0$, which gives the universality class of the discrete model one starts with. After recalling the construction of classical SLE, the goal of the presentation will be to explain how a similar line of thinking can be used to introduce local multiple SLE measures and to obtain a similar classification (local because it turns out this approach only allows to describe initial segments of the curves). This measures are used to describe simultaneously the scaling limit of multiple interfaces of critical 2D statistical mechanics models. The results I will present are due to Dubédat in 2006 and 2007 later developed from 2015 to a few years ago by other groups of mathematicians (K. Kytölä, A. Karrila, E. Peltola and others). I will only speak about the chordal case and I will not go into any details of the proofs but only give briefly the ideas. If times allows I will briefly explain how this topic is related what I am currently attempting to do in my own research.

Salle : P205

Rencontres statistiques

---

ZAFFRAN Margaux (Université Paris-Saclay)

Le 23/03/2026
De 13:45 à 14:45
Titre : On the hardness of group-conditional distribution-free predictive inference, an application to prediction with missing covariates.

Résumé : Predictive uncertainty quantification is crucial in decision-making problems.In this talk, we will focus on distribution-free uncertainty quantification by considering predictive intervals for the target Y enjoying validity (i.e. nominal coverage) with no assumptions on the underlying data generating process nor the sample size. After introducing the framework, we will detail the nuance between marginal validity and conditional---on the test point---validity. We will review the existing (impossibility) results on conditional validity. This will lead us to our main question: how can we relax the goal of conditional validity to make it achievable? We will present new hardness results, that characterize the limits of group conditional coverage (e.g., achieving nominal coverage not only on average but also among women on the one hand, and among men on the other hand), a weaker goal extensively used in the literature in place of the impossible perfect conditional validity. Finally, we will dive into applying these results in the context of prediction with missing values. There, one wants to obtain not only marginally valid intervals despite missing values, but also intervals that achieve the nominal coverage regardless of which values are missing at test time. We provide an algorithm reaching this goal by constructing informative predictive intervals in light of our hardness results. Based on a joint work with J. Josse, Y. Romano & A. Dieuleveut

Salle : A307

Séminaire Analyse-Probabilités

---

MARÊCHÉ Laure (Université de Strasbourg)

Le 24/03/2026
De 10:30 à 11:30
Titre : A general limit theorem for self-interacting random walks

Résumé : In this talk, we will present self-interacting random walks, a class of non-Markovian random walks, such that the probability the walk goes to a given location is smaller if, in the past, it has often crossed the edge between its current position and the target. Tóth introduced these processes in the 90s, and investigated limit theorems for their trajectory. However, he could not prove such results, and even nowadays only a few cases are proven. We will present a general limit theorem potentially applicable to all cases, and use it to prove one of the major remaining cases.

Salle : A711

Groupe de travail Probabilites

---

BOUIN Emeric ()

Le 01/04/2026
De 14:00 à 15:15
Titre : TBC

Résumé :

Salle : P205

Colloquium du CEREMADE

---

DESVILLETTES Laurent (Université Paris Cité)

Le 07/04/2026
De 15:30 à 16:30
Titre : A PDE approach for some theoretical problems in population dynamics

Résumé : We consider integrodifferential equations for populations which are structured with respect to a quantitative trait, and which takes into account the phenomena of selection, mutations, and competition between individuals. We explain the richness of the associated dynamics, and the variety of large time behaviors of the solutions. We also present some interpretations in population dynamics and evolution theory. The presentation is mainly based on works in collaboration with Angel Calsina, Silvia Cuadrado, Pierre-Emmanuel Jabin, Stéphane Mischler and Gaël Raoul.

Salle : A709

Séminaire Analyse-Probabilités

---

SUN Changzhen (CNRS & Laboratoire de Mathématiques de Besançon)

Le 14/04/2026
De 10:30 à 11:30
Titre :

Résumé :

Salle : A711

Groupe de travail Probabilites

---

BLACHAR Guy ()

Le 15/04/2026
De 14:00 à 15:15
Titre : TBC

Résumé :

Salle : B126

Groupe de travail Probabilites

---

TBC ()

Le 07/05/2026
De 14:00 à 15:15
Titre : TBC (Attention Thursday)

Résumé :

Salle : D205

Séminaire Analyse-Probabilités

---

AYI Nathalie (LJLL)

Le 19/05/2026
De 10:30 à 11:30
Titre : TBA

Résumé : TBA

Salle : A711

Groupe de travail Probabilites

---

TBC ()

Le 20/05/2026
De 14:00 à 15:15
Titre : TBC

Résumé :

Salle : D205

Groupe de travail Probabilites

---

MARINI Elisa ()

Le 03/06/2026
De 14:00 à 15:15
Titre : TBC

Résumé :

Salle : D205

Groupe de travail Probabilites

---

TBC ()

Le 17/06/2026
De 14:00 à 15:15
Titre : TBC

Résumé :

Salle : D205