Speaker
Márcio Ferreira
(University of Coimbra)
Description
Two recent applications of generative machine learning models to neutron star physics will be presented.
i) An anomaly detection framework based on normalizing flows (NF) models to detect the presence of a large (destabilizing) dense matter phase transition in neutron star (NS) observations of masses and radii, and relate the feasibility of detection with parameters of the underlying mass-radius sequence. The NF models allow to determine the likelihood of a first-order phase transition in a given set of M(R) observations featuring a discontinuity.
ii) An inference framework based on conditional variational autoencoders to reconstruct the neutron star equation of state from a given set of mass-radius observations will be presented.
Primary author
Márcio Ferreira
(University of Coimbra)
