Title: Unveiling the Universe with Gravitational Waves: a Machine Learning approach

Venue & Time: Blue Room / 10:30

Abstract: This thesis is devoted to the study of gravitational waves, combining theoretical investigations with novel applications of machine learning. The first part of the work addresses the dynamics of hyperbolic encounters, a class of transient gravitational wave signals. By incorporating orbit precession and post-Newtonian corrections, this study provides theoretical insights into the emission properties.

The second part shifts toward data-driven approaches, exploring the use of advanced machine learning methods in gravitational waves field. Two main applications are considered: constraining the energy density of the stochastic gravitational wave background through astrometric measurements and inferring the properties of lensed gravitational wave signals. In both cases, the machine learning frameworks developed here show significant gains in computational efficiency, with particular promise for real-time signal analysis and alerts, demonstrating the transformative role of machine learning in enhancing current and future gravitational wave data analysis strategies.

Supervisors: Savvas Nesseris and Sachiko Kuroyanagi