Choose timezone
Your profile timezone:
Description
The physics programme at the LHC and future colliders is increasingly bottlenecked by the computational cost and flexibility of the classical analysis pipeline. Machine learning can help overcome these limitations, but must be physics-informed to ensure accuracy, flexibility, and robust generalization. I will first give an overview of machine learning applications at the LHC, then present examples spanning QFT amplitude surrogates, jet taggers, and event-level generative models, with an emphasis on Lorentz-equivariant networks.
