Machine Learning at the Flatiron Institute Seminar: Francois Lanusse

Title: Merging Deep Learning with Physical Models for the Analysis of Cosmological Galaxy Surveys

Abstract: As we move towards the next generation of galaxy surveys, our field is facing outstanding challenges at all levels of scientific analysis, from pixel-level data reduction to cosmological inference. As powerful as deep learning has proven to be in recent years, in most cases deep learning 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 its integration in scientific analyses.

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 practice this will mean following two main directions: 1. using deep generative models as a practical way to manipulate implicit distributions (either data- or simulation-driven) within a larger Bayesian framework. 2. Developing, at scale, automatically differentiable physical models amenable to gradient-based optimization and inference.

I will illustrate these concepts in a range of applications in the context of cosmological galaxy surveys, from pixel-level astronomical data processing (e.g. deconvolution), to inferring cosmological parameters through fast and automatically differentiable cosmological N-body simulations. Methodology-wise these examples will involve in particular diffusion generative models, score-enhanced simulation-based inference, and hybrid physical/neural ODEs.