Gravitational data analysis with machine learning and deep learning

Gravitational-wave astronomy enables precision studies of compact binary systems and provides powerful tests of General Relativity. Black-hole spectroscopy—the detection of gravitational-wave emission spectra from black-hole ringdowns—offers a particularly clean framework for testing gravity theories against well-defined predictions.

However, accurate waveform modeling and efficient parameter estimation are essential to extract both astrophysical and fundamental physics insights from the data. In this talk, I will present my recent works in gravitational waveform modeling and inference, with a focus on black hole ringdowns, and highlighting the role of machine learning and deep learning techniques.

In particular, I will discuss the promises and challenges of using Gaussian Process Regression for waveform modeling and hierarchical inference, and simulation-based inference for parameter estimation.