PyTorch Basics: Machine Learning for Physics

Thursday 9 Jul 2026, 14:00 → 16:00 Europe/Zurich

593/R-010 - Salle 11 (CERN)

Abstract

During a practical session on Deep Learning, we will build, train, and evaluate a Neural Network from scratch using PyTorch to solve a classic high-energy physics problem: classifying signal (electrons/photons) from background (hadronic jets).

Lecturer Bio

Davide Valsecchi is a Postdoctoral Researcher at ETH Zurich, based at CERN with the CMS Experiment. He has been working at the intersection of particle physics and AI since his PhD, where he built one of the first Graph Neural Networks for CMS electron and photon clustering. Currently, Davide coordinates the CMS E/Gamma Physics Object Group and recently led the effort to integrate PyTorch into the CMS software stack. His active research focuses heavily on Generative Deep Learning—specifically using Normalizing Flows for unbinned likelihood fits, unfolding and detector calibration.