Kimia Nadjahi, a second year doctoral student at Télécom Paris and a member of the DSAIDIS Chair, has won the best student paper award at ICASSP 2020 on 5 May.  Her work was singled out for its innovation, impact and quality, as well as for her oral presentation at this research conference, which focuses on signal processing and machine learning.

The 45th ICASSP conference was held entirely online in May 2020. A committee of experts selected Kimia Nadjahi, whose main research areas are inferential statistics and optimal transport, as the recipient of the  best student paper award. Kimia’s research is supervised by Umut Şimşekli and Roland Badeau at Télécom Paris. Her work aims to detect the way in which practical methods for optimal transport, inferential statistics and implicit generative modeling can be combined, to solve large-scale machine learning problems.

The title of the award-winning article is:

Approximate Bayesian Computation Method with the Sliced-Wasserstein Distance, by Kimia Nadjahi, Valentin De Bortoli, Alain Durmus, Roland Badeau, Umut Şimşekli.

Read the article in the HAL database