Neural Networks unveiling the properties of gravitational wave background from massive black hole binaries

Abstract

Massive black hole binaries (MBHBs) are binary systems formed by black holes with mass exceeding millions of solar masses, expected to form and evolve in the nuclei of galaxies. The extreme compact nature of such objects determines a loud and efficient emission of Gravitational Waves (GWs), which can be detected by the Pulsar Timing Array (PTA) experiment in the form of a Gravitational Wave Background (GWB), i.e. a superposition of GW signals coming from different sources. The modelling of the GWB requires some assumptions on the binary population and the exploration of the whole involved parameter space is prohibitive as it is computationally expensive. We here train a Neural Network (NN) model on a semi-analytical modelling of the GWB generated by an eccentric population of MBHBs that interact with the stellar environment. We then use the NN to predict the characteristics of the GW signal in regions of the parameter space that we did not sample analytically. The developed framework allows us to quickly predict the level, shape and variance of the GWB signals produced in different universe realisations.