From Quarks to Neural Networks: Mapping the Proton with Machine Learning

by Prof. Emanuele Roberto Nocera (University of Turin)

Wednesday Nov 19, 2025, 3:00 PM → 4:00 PM Europe/Rome

Dipartimento/1-1 - Sala Consiglio (Dipartimento)

The proton, one of the fundamental building blocks of visible matter, conceals a complex structure in which quarks and gluons interact through the strong force described by Quantum Chromodynamics (QCD). Revealing this structure is essential for understanding high-energy particle scattering, as can be studied at colliders, for testing the boundaries of the Standard Model, and for discovering New Physics signatures. In this colloquium, I will show how we can reconstruct the proton’s internal landscape through global analyses of parton distribution functions (PDFs), which connect collider data to the underlying dynamics of QCD. I will specifically focus on how modern machine learning techniques - neural networks, stochastic optimization, and robust statistical validation - allow us to extract this information with unprecedented precision and accuracy.
The resulting picture exemplifies how physics and artificial intelligence can work together to deepen our understanding of nature at its most fundamental scale.