About me
Reseach interests: optimization, distributed/federated learning, low-cost training, clients statistical heterogeneity, unsupervised learning.
Ph.D. defense: You can find my manuscript here and also my slides!
MT180s: I participated to MT180s (equivalent of 3MT) and reached the regional final. Here is the video!
I am a postdoctoral researcher at DI ENS and Inria Paris working with Kevin Scaman and Laurent Massoulié.
I obtained my Ph.D. in “Applied Mathematics and Computer Science” from École Polytechnique (Paris) under the joint supervision of Eric Moulines and Aymeric Dieuleveut, professors at the CMAP laboratory of École polytechnique.
The goal of the thesis is to focus simultaneously on two challenges of federated learning (i.e. machine learning distributed across millions of devices):
- reducing the cost of communication by doing bidirectional compression,
- while considering a heterogeneous setting.
I graduated from Ensimag (Grenoble) with a specialisation in computer science, applied mathematics and 3D animation.
I also had the opportunity to do a six-month academic exchange with the MIPT (Moscow) to study statistics, machine learning and image recognition under the supervision of Vadim Strijov.