Welcome!

I am currently a postdoctoral researcher in Optimix team in LIX (Laboratoire d’Informatique de l’Ecole Polytechnique) where I work with Catuscia Palamidessi and Sonia Vanier, in a partnership with Crédit Agricole (French bank). I work on data privacy in machine learning, especially differential privacy. I am especially interested in the theoretical aspects of privacy, but also in its relations with other ethical properties like frugality, robustness, fairness.

Previously, I did my PhD between 2020 and 2023 in CEA List under the direction of Renaud Sirdey and the cosupervision of Cédric Gouy-Pailler. We proposed approaches that combine differential privacy and homomorphic encryption to protect the training data privacy in collaborative machine learning (federated learning, PATE) from any stakeholder (server, end-users, other data owners).

Before my PhD, I worked as a research engineer in LIP6 in 2019, under the supervision of Nicolas Maudet, Patrice Perny and Paolo Viappiani, in the area of computational social choice. We proposed and studied a voting rule suited for the situations where some candidates may turn out to be unavailable after the vote.

Before this, I entered research working in CEA List for C-BORD project, under the supervision of Jean-Philippe Poli, between 2017 and 2018. I designed an explainable classification algorithm dealing with imprecise inputs thanks to fuzzy decision trees.

Main research interests

• data privacy, in particular differential privacy
• collaborative machine learning
• frugal machine learning
• computational social choice