Speaker
Pablo Villanueva
Description
Machine learning provides us with a powerful tool in the fields of cosmology and astrophysics, allowing us to better understand how matter clusters and is distributed in the universe. In this talk, we review some relevant cosmological applications, focusing on two different methods. On the one hand, extracting the matter density field from 21 cm hydrogen maps making use of Convolutional Neural Networks (CNNs). On the other hand, inferring the total mass of a dark matter halo from a few properties of its hosted galaxies via Graph Neural Networks (GNNs). This provides a novel method to weigh real halos such as the Milky Way and Andromeda.
