Francois Lanusse

CNRS/CEA Paris-Saclay

ABSTRACT: As we move towards the next generation of cosmological surveys, our field is facing new and outstanding challenges at all levels of scientific analysis, from pixel-level data reduction to cosmological inference. As powerful as Deep Learning (DL) has proven to be in recent years, in most cases a DL approach alone proves to be insufficient to meet these challenges, and is typically plagued by issues including robustness to covariate shifts, interpretability, and proper uncertainty quantification, impeding their exploitation in scientific analysis.

In this talk, I will instead advocate for a unified approach merging the robustness and interpretability of physical models, the proper uncertainty quantification provided by a Bayesian framework, and the inference methodologies and computational frameworks brought about by the Deep Learning revolution.

In particular, we will see examples of this approach at various stages of the cosmological analysis of upcoming wide-field survey, with applications ranging from   the analysis of individual galaxy images, to inferring cosmological parameters through hybrid physical/deep learning models.