Program 2022

 

StatMathAppli 2022 - Sunday 28 August to Friday 2 September - La Villa Clythia, Fréjus

 

The plenary room is the DESCARTES room


Sunday 28 August

18:00 - 22:00 : Welcome and dinner from 19:30 until 21:00 (cold buffet in self-service)

 

Monday 29 August

08:45 - 9:00 : Welcome and delivery of badges

9:00 - 11:00 : Lecture 1 by Richard Nickl

11:00 - 11:30 : Break

11:30 - 12:30 : Contributed talks, session 1: Bayesian inference and generative models

Geerten Koers: Bayesian linear methods for non-linear inverse problems.

Clément Berenfeld: Bayesian density estimation near unknown manifolds.

Jakiw Pidstrihgach: Score-based generative models detect manifolds.

12:30 - 15:30 : Lunch

15:30 - 17:30 : Lecture 2 by Rebecca Willett

17:30 - 18:00 : Break

18:00 - 19:00 : Contributed talks, session 2: Machine learning and optimisation

Julian Tachella: Equivariant imaging: learning to solve imaging inverse problems without ground-truth.

Constantin Philippenko: Preserved central model for faster bidirectional compression in distributed settings.

Vincent Divol: Estimation of optimal transport maps in high-dimension.

19:30 - 21:30 : Friendly aperitif and Dinner

 

Tuesday 30 August

09:00 - 11:00 : Lecture 3 by Richard Nickl

11:00 - 11:30 : Break

11:30 - 12:30 : Contributed talks, session 3: Network inference and clustering

Emre Anakok: Accounting for the observational process when unraveling the structure of the interaction network.

Camille Champion: Microbial network inferences and clustering.

Deborah Sulem: Regularized spectral methods for clustering signed networks.

12:30 - 15:30 : Lunch

15:30 - 16:00 : Flash talks (2 minutes per speaker)

16:00 - 17:00 : Poster Discussion

Bernhard Stankewitz: Sequential early stopping for L2-boosting in high-dimensional linear models.

Guillaume Maillard: Robust estimation in total variation distance under a shape constraint.

Maté Kormos: Caliper matching estimator of average treatment effects.

Olympio Hacquard: Classification of measures, application to persistence diagrams.

Jean-Baptiste Fermanian: Stein’s phenomenon for the multi-task averaging problem in high-dimension.

Eddy Ella Minsta: Classification procedure for diffusion paths.

Linus Bleistein: Theoretical Guarantees for learning with signatures.

Maia Tienstra: Frequentist Ensemble Kalman filter.

Diksha Bhandari: Neural network modelling of brain responses during language comprehension.

17:00 - 17:30 : Break

17:30 - 19:00 : Lecture 4 by Rebecca Willett

19:30 - 21:30 : Provencal Dinner

 

Wednesday 31 August

09:00 - 10:30 : Lecture 5 by Richard Nickl

10:30 - 11:00 : Break

11:00 - 12:30 : Lecture 6 by Rebecca Willett

12:30 - 15:00 : Lunch

15:00 - 19:30 : Free time / Activities

19:30 - 21:30 : Dinner

 

Thursday 1 September

09:00 - 11:00 : Lecture 7 by Richard Nickl

11:00 - 11:30 : Break

11:30 - 12:30 : Contributed talks, session 4: Adaptive inference and testing

Perrine Lacroix: Penalty functions calibration for high-dimensional Gaussian linear regression.

Tom Guédon: Variance components testing in mixed effects models in small sample size.

Neil Deo: On Adaptive confidence sets for the Wasserstein distances.

12:30 - 16:00 : Lunch

16:00 - 17:30 : Lecture 8 by Rebecca Willett

17:30 - 18:00 : Break

18:00 - 19:00 : Contributed talks, session 5: Bandit problems

Dorian Baudry: Optimal Thompson Sampling strategies for support-aware CVaR bandits.

Marc Jourdan: Top two algorithms revisited.

El Mehdi Saad: Constant regret for sequence prediction with limited advice.

19:30 - 21:30 : Dinner

 

Friday 2 September

09:00 - 10:30 : Lecture 9 by Rebecca Willett

10:30 - 11:00 : Break

11:00 - 12:00 : Contributed talks, session 6: Temporal data analysis

Grégoire Szymanski: Optimal estimation of the rough Hurst parameter.

Liudmila Pishchagina: Multiple change-point detection problem for bi-variate time-series.

Alexandre Lecestre: Robust estimation in finite state space hidden Markov models.

12:00- 14:00 : Lunch and leaving