Constraining Neutron Star EoS Using Machine Learning and the CompOSE Database

Astronomical data collected so far has not been sufficient to determine the internal structure of neutron stars (NS). One of the main telltales of the internal structure of a NS is its equation of state (EoS), i.e., the relation between its internal pressure and density. The NS EoS has been studied theoretically using numerous phenomenological and microscopic nuclear models. Many of these EoS have very different underlying assumptions. Comparison of such models with observations have not been able to provide compelling evidence so far of the best theoretical model found in the literature. As a result, there are many different EoS that reproduce astronomical observations equally well. To further constrain and find the most suitable EoS, one has increasingly adopted Bayesian analysis techniques. This method has been undertaken by many authors to probe the impact of phenomenological and microscopic EoS on known data, including recent observations from NICER and LIGO. We are currently working with this method using applications of Machine Learning (ML) to improve the constraints placed on the several EoS, comparing them with astronomical observations and with available nuclear reaction data. We extend our group's previous works on ML by including numerous EoS compiled by the CompOSE task force.