2:00–2:30 pm ERC 445
François Lanusse (CNRS) "Merging Deep Learning with Physical Models for the Analysis of Cosmological Surveys"
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 how deep generative models can be embedded within principled physical Bayesian modeling to solve a number of astronomical ill-posed inverse problems ranging from deblending galaxy images, all the way to inferring the distribution of dark matter from weak gravitational lensing measurements. I will also illustrate how the same generative modeling techniques can alleviate the need for analytic likelihoods in cosmological inference, enabling instead Simulation-Based Inference in which the physical model is implemented in the form of a numerical simulator. And finally, I will highlight the power of the computational frameworks initially developed for Deep Learning when applied to physical modeling, with applications ranging from speeding up cosmological MCMCs to performing inference over the initial conditions of N-body simulations.
