Cosmology is a study to understand the origin, fundamental property, and evolution of the universe. Nowadays, many observational data of galaxies have become available, and one needs large-volume numerical simulations with good quality of the spatial distribution for a fair comparison with observation data. On the other hand, since galaxies' evolution is affected by both gravitational and baryonic effects, it is nontrivial to populate galaxies only by N-body simulations. However, full hydrodynamic simulations with large volume are computationally costly. Several attempts have been proposed to overcome such difficulties, such as applying alternative galaxy assignment methods or painting baryonic feature learned from hydrodynamic simulations to N-body simulations.
In this talk, I would like to introduce the MBP-galaxy abundance matching and mock galaxy catalogs of the Horizon Run 4 and Multiverse simulations, large-volume cosmological N-body simulations done by the Korean community. Also, I would like to briefly discuss how recent deep-learning techniques could help cosmological studies.