A machine learning approach to mapping baryons on to dark matter haloes using the <scp>eagle</scp> and <scp>C-EAGLE</scp> simulations

ABSTRACT High-resolution cosmological hydrodynamic simulations are currently limited to relatively small volumes due their computational expense. However, much larger required probe rare, overdense environments, and measure clustering statistics of the large-scale structure. Typically, zoom individual regions used study rare semi-analytic models halo occupation applied dark-matter-only (DMO) Universe in large-volume regime. We propose a new approach, using machine learning framework, explore halo–galaxy relationship periodic eagle simulations, C-EAGLE galaxy clusters. train tree-based method predict baryonic properties galaxies based on host dark matter properties. The trained model successfully reproduces number key distribution functions for an infinitesimal fraction cost full simulation. By training both zooms we learn bias evolution differing environments. This allows us apply DMO volume than would be possible if only demonstrate this application (800 Mpc)3 P-Millennium simulation, present predictions from large volume.