Fabien Pesquerel

Postdoctoral researcher in the PIRAT\'); team researching on Reinforcement Learning x Cybersecurity.

profile_picture.jpg

INRIA, CentraleSupelec Rennes

Cesson-Sevigné, FRANCE

I am interested in the problem of sequential learning in which an agent interacts with an environment in order to achieve a goal. The agent must adapt to the information obtained from its observations. The sequential processing of information make this problem a difficult and interesting one, in particular when applied to machine translation, recommender systems, physics, or even cybersecurity.

What makes such problems interesting to me is that you learn about the problem you want to solve as you solve it!

I am as motivated by the mathematical and computational aspects of the problem (What is an optimal strategy/decision?) than I am by the modelling aspects (How can we model a human decision-making? How to model and incroporate expert knowledge?).

Outside of work, I really enjoy cooking and doing calisthenics. I have a blog AnalySthenics that is mainly dedicated to the analysis of calisthenics and gymnastics data.

news

Oct 22, 2024 :japan: I am starting to learn Japanese. Hope to progress steadily and quickly :mount_fuji:!
Apr 01, 2024 :office: I am starting a postdoc in the PIRAT team at INRIA, CentraleSupelec Rennes!
Dec 25, 2023 It is Christmas :christmas_tree: and I recently sucessfully defended my PhD :mortar_board: ! Read the manuscript :book: here!

selected publications

  1. IMED-RL: Regret optimal learning of ergodic Markov decision processes
    Fabien Pesquerel, and Odalric-Ambrym Maillard
    In Advances in Neural Information Processing Systems, 2022
  2. PhD thesis
    Information per unit of interaction in stochastic sequential decision making
    Fabien Pesquerel
    Dec 2023