Gravitational wave data from LIGO, Virgo and KAGRA typically require a lot of computational power to be analyzed. Traditional methods like matched filtering for the search or bayesian inference for the parameter estimation are very accurate and reliable, but slow. In this seminar I will present some novel techniques that are being introduced in the field based on machine learning. Particularly, I will focus on the detection of compact binary coalescences and on the determination of its parameters. The results obtained show that using such techniques the computing time is reduced by several orders of magnitude, yet the performance is not yet equal to that of traditional methods.
Giada Caneva, Elia Bertoldo, Clara Fernandez Castañer