Julien Aubert

PhD student at the Université Côte d'Azur

Julien Aubert

About Me

I am a PhD student in Applied Mathematics at the Université Côte d'Azur, where I am part of the J. A. Dieudonné laboratory. I am currently working on parameters estimation in bandit algorithms under the supervision of Patricia Reynaud-Bouret and Luc Lehéricy. The purpose of our work is to establish a theoretical framework for fitting parameters of cognitive models in neurosciences. Prior to my PhD, I studied at the Ecole Centrale-Supélec and also obtained an MSc in Mathematical Sciences from Oxford University.

Research Interests

My PhD research focuses on modeling learning processes based on individual behavioral data. I am investigating the theoretical properties of maximum likelihood estimation in the context of non-stationary and dependent data. This involves identifying the model that best captures an individual's behavior within a specified model class. Key topics include concentration inequalities, penalized maximum likelihood estimation, and hold-out procedures. The models we use are reinforcement learning type models, and more specifically contextual bandits models. Additionally, I explore applications in ethology and categorization tasks.

Publications

Learning how contextual bandits learn

Preprint, 2024

Read Paper
General oracle inequalities for a penalized log-likelihood criterion

Preprint, 2024

Read Paper
Exponential inequalities for suprema of processes with stochastic normalization

Preprint, 2024

Read Paper
On the convergence of the MLE as an estimator of the learning rate

ICML, 2023

Read Paper

Teaching Experience

Curriculum Vitae

Download my CV

Contact Me

You can reach me via email at:

julien.aubert@univ-cotedazur.fr

Social links