Provisional program 2025

The program can be downloaded here.

The plenary room is the DESCARTES room.



Sunday, August 31th

18.00 - 21:00 : Welcome and dinner from 19:30 until 21:00 (cold buffet in self-service)


Monday, September 1st

08.45 - 9:00 : Welcome and delivery of badges

09:00 - 10:30 : Lecture 1 by John Duchi 

10:30 - 11:00 : Break

11:00 - 12:20 : Contributed talks Privacy and Robustness

  • Alexander Kent. Rate optimality and phase transition for user-level local differential privacy.
  • Màté Kormos. Private double robust inference.
  • Thomas Michel. DP-SPRT: differentially private sequential probability ratio tests.
  • Richard Schwank. Robust score matching

12:30 -13.30 : Lunch

15:30 - 17:00 : Lecture 1 by Po-Ling Loh

17:00 - 17:30 : Break

17:30 - 19:00 : Invited talk "Some latent variable models in ecology" by Stéphane Robin.

19:15 : Friendly aperitif and Dinner


Tuesday, September 2nd

09:00 - 10:30 : Lecture 2 by Po-Ling Loh 

10:30 - 11:00 : Break

11:00 - 12:30 : Lecture 2 by John Duchi

12:30) 13.30 : Lunch

15:30 - 17:00 : Invited talk "Interplay between data, physics and simulation models with a link to industrial use-cases" by Mathilde Mougeot.

17:00 - 17:30 : Break

17:30 - 18:00: Flash talks (2 minutes per speaker)

  • Alberto Bordino: Density ratio permutation tests with connections to distributional
    shifts and conditional two-sample testing
  • Jean-Baptiste Fermanian: Uncertainty reduction of class conditional conformal
    prediction via multi-inputs aggregation
  • Maximilian Graf: Statistical inference for paired spatial Poisson processes with
    missing data
  • Harold Guéneau: Money laundering detection: financial time series representation
    learning with a transformer by contrastive learning
  • Anton Kutsenko: Complete tail asymptotics for branching processes
  • Félix Laplante: Uniform nonparametric confidence bands for random cumulative
    distribution functions
  • Romain Périer: Post hoc bounds for heterogeneous data
  • Sylvain Procope-Mamert: Iterative forward scheme to construct proposals for
    sequential Monte Carlo Algorithms
  • Paul Rosa: On L^2-posterior contraction rates in Bayesian nonparametric regression
    models
  • Vincent Runge: DUST; a duality-based pruning method for exact multiple
    change-point detection
  • Henning Stein: Gentle measurements of quantum states
  • William Underwood: Model upgrading in survival analysis

18:15 - 19:15: Poster Discussion 

19:30 : Provencal Dinner


Wednesday, September 3rd

09:00 - 10:30 : Lecture 3 by John Duchi

10:30 - 11:00 : Break

11:00 - 12:30 : Lecture 3 by Po-Ling Loh

12:30 -13:30 : Lunch

14:30 - 19:30 : Free time / Activities

19:30 : Dinner


Thursday, September 4th

09:00 - 10:30 : Lecture 4 by Po-Ling Loh

10:30 - 11:00 : Break

11:00 - 12:20 : Contributed talks Robustness and optimisation

  • Renaud Gaucher. A unified breakdown analysis for Byzantine robust Gossip.
  • Laurentiu Marchis. On the benefits of accelerated optimization in robust and private estimation.
  • El Mehdi Saad. New lower bounds for stochastic non-convex optimization through divergence decomposition.
  • Laura Hucker. Comparing regularisation paths of (conjugate) gradient estimators in ridge regression.

12:30 - 13.30 : Lunch

15:30 - 17:00 : Lecture 4 by John Duchi

17:00 - 17:30 : Break

17:30 - 18:50 : Contributed talks Applied statistics

  • Matthieu Texier. Combining mixture models and Markov chains to explore spatio-temporal dynamics of child wasting in southern Madagascar.
  • Tom Rohmer. Improving genetic parameter estimation for dependent traits under selection.
  • Marina Gomtsyan. Variable selection in specific regression for count time series
  • Christophe Ley. A versatile trivariate wrapped Cauchy copula with applications to toroidal and cylindrical data.

19:30 : Dinner


Friday, September 5th

09:00 - 10:20 : Contributed talks Unsupervised learning

  • Bertrand Even. Computational lower bounds for latent variables: clustering, sparse clustering and biclustering.
  • Victor Thuot. Clustering items through bandit feedback: finding the right feature out of many
  • Ibrahim Kaddouri. Clustering in slowly mixing Gaussian hidden Markov models.

10:20 - 10:40 : Break

10:40 - 11:40 : Contributed talks Statistical inference

  • Paul Rognon-Vael. Improving variable selection properties by using external data.
  • Sophia Loizidou. Optimal tests for symmetry on the torus.
  • Philippe Berthet. Some recent applications of Gaussian couplings in empirical process theory.

12:00 : Lunch and leaving