I am a post-doc working on adversarial images, kernel methods, reinforcement learning and causality.

I did my PhD at the Max Planck Institute for Intelligent Systems in Tübingen (Germany), under the supervision of Prof. Bernhard Schölkopf. My PhD thesis “Distribution-Dissimilarities in Machine Learning”  presents many well-known distribution dissimilarities (KL- and f-divergences, total variation, Wasserstein-distances, kernel mean embeddings) as specific types of “classifier based dissimilarities”, which happen to be implicitly used in state-of-the-art generative algorithms such as  GANs and VAEs. At the beginning of my PhD, I also briefly worked on causality applied to exoplanet detection, as well as bio-computational algorithms for genomic wide association tests (GWAS).

I am also very interested in energy and climate change related questions, and hope to be able, one day, to efficiently help reducing our carbon footprint.