cv
General Information
| Full Name | Fabien Pesquerel |
| Languages | French, English |
Education
-
2020 - ongoing
PhD
Institut National de Recherche en Informatique et en Automatique (INRIA), Villeneuve d'Ascq, France
- I do research in Reinforcement Learning and am interested in exploiting structures in sequential decision making tasks.
- This PhD thesis is made possible thanks to a grant from the École Normale Supérieure de Paris (ENS).
-
2016 - 2020
Master's degrees of Computer Science and Applied Mathematics
École Normale Supérieure de Paris (ENS), Paris, France
-
2015
Bachelor's degree of Physics
École Normale Supérieure de Paris (ENS), Paris, France
Experience
-
2021 - 2022
Teacher Assistant
École Polytechnique, École Centrale-Supélec, Université de Lille
- Reinforcement Learning courses at the graduate level.
- Organized and taught practical sessions.
-
2020 - 2021
Teacher Assistant
École Polytechnique, École Centrale-Supélec
- Reinforcement Learning courses at the graduate level.
- Organized and taught practical sessions.
-
2020
Intern
Institut National de Recherche en Informatique et en Automatique (INRIA), Villeneuve d'Ascq, France
- I worked on bandits problems with group-like structures in the SequeL team with Odalric-Ambrym Maillard.
-
2019
Intern
École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland
- I worked on 3D model generation in the Computer Vision Laboratory (CVLab) with Edoardo Remelli and Pascal Fua.
-
2019
Intern
Université Jean Monnet, Saint-Étienne, France
- I worked on metric learning problem with highly imbalanced data in the Data Intelligence team of the Hubert Curien Laboratory with Léo Gautheron, Marc Sebban and Amaury Habrard.
Open Source Projects
-
2022 - ongoing
ComputationalStatistics_Class
- A numerical tour of Computational Statistics.
-
2022 - ongoing
ReinforcementLearning_Class
- A numerical tour of Reinforcement Learning and sequential decision making.
Academic Interests
-
Statistics & Optimization
- In both statistics and optimization, I seek tools to scrutinize the world of world's model. I want to build theoretical tools to know what we want to know and to do it as fast as possible. It raises the ill posed question; what is the optimal rate of information acquisition on a given problem we want to solve?
- What is a good model? When is a model better than another? How to take into account some uncertainty? Those are some questions that I am interested in.
-
Sequential Decision Making
- In between statistics and optimization, I like to seek how to exploit a stream of information to optimize a metric in an online fashion.
- Given descriptive models of the world, what is the best next experiment I can make to decide on the best model? Given a state of knowledge, is there a logical way to maximally improve my future state of knowledge? Whether or not those are good questions, I am interested in thinking about them through the sequential decision making lens.
Other Interests
- I am a calisthenics and artistic gymnastics enthusiast. I recently took a liking into bouldering and rock climbing in general. I love walking!