Speaker
Description
Hadron resonances are rigorously defined as poles of scattering amplitudes in the complex energy plane. Experimentally, however, we only have access to data along the real-energy axis, and even there, measurements are limited to observables such as cross sections, which are only sensitive to the squared modulus of the underlying complex amplitude. As a result, reconstructing the full scattering amplitude and extracting resonance parameters from noisy data is a highly ill-posed inverse problem. In this talk, I will review recent efforts to address these challenges using machine-learning techniques. In particular, I will focus on ongoing work that leverages physics-informed generative models to reconstruct scattering amplitudes in a way that respects unitarity and other physical constraints.
