I am a data scientist at Owkin, a French AI biotech company. Prior to that, I was a post-doctoral researcher at the Sierra INRIA team. I defended my PhD in October 2020 at ENSAE, where I worked under the supervision of Marco Cuturi.
Within Owkin, my research lies at the intersection of federated learning and bioinformatics. During my PhD, I leveraged the closed form of OT between elliptical distributions, along with regularized approaches of OT, to make a few contributions towards helping OT theory overcome some of its computational and statistical drawbacks, and gain applicability in machine learning. Starting from my postdoc, I got interested in using kernel methods to design algorithms that alleviate the curse of dimensionality in OT between distributions with smooth densities.
- [15/12/2022] Just released PyDESeq2, a python package for bulk RNA-seq differential analysis.
- [29/11/2022] I co-presented two papers from our team at NeurIPS: SecureFedYJ: a safe feature Gaussianization protocol for Federated Learning and FLamby: Datasets and Benchmarks for Cross-Silo Federated Learning in Realistic Healthcare Settings.