Neutron star mergers: numerical models to nuclear-physics inference

In this Thesis, we explore how neutron star mergers, the collisions of dead stars, can be used to further our understanding of the physics of extremely dense matter. First, we develop a new code that improves the simulation of the material thrown out into space by these collisions. Secondly, we build data-analysis and machine-learning tools to better analyse the gravitational-wave signals from these mergers. We show that our methods improve our understanding of nuclear physics and cosmic expansion.