Current gravitational waves observatories detect signals from the coalescence of binary systems of black holes and/or neutron stars. For such events, the most accurate solutions of Einstein Equations are obtained with numerical relativity simulations, that are not available in sufficient quantity to cover the full space of parameters describing astrophysical objects. Parameter estimation therefore relies on templates obtained from approximate methods, generally described as "gravitational waveform models”. Recently, applications of machine learning algorithms enabled to increase the accuracy of gravitational waves templates, notably for precessing systems described by a large parameter space. This talk will review the state of the art in models and describe the advantages presented by neural networks and non-parametric modelling, showcasing their ability to provide precise and feasible parameter estimation for the next generation of instruments.
Copyright. All Rights Reserved.