Exploring the constraints on the equation of state from neutron star macroscopic properties using neural network algorithms

Given the large uncertainty in the equation of state (EOS) for dense nuclear matter and the difficulty of reproducing it in terrestrial laboratories, the development of astronomical observations has provided important insights. The macroscopic properties of neutron stars provide an important basis for the inverse constrained EOS of dense nuclear matter. Currently, several methods have been proposed to implement this inverse constraint, including the Bayesian analysis, the Lindblom’s approach, and the polynomial regression, and so on. In recent years, neural network algorithms have demonstrated remarkable capabilities in solving complex inverse problems and have gradually been applied in the neutron star research. Based on this algorithm, we use an isospin-dependent parameterized EOS model to construct a neural network that maps from several data of observable properties to the EOS. Moreover, the trained neural network’s prediction of the EOS can be further validated against results from other studies.