Orateur
Stefano Forte
(Dipartimento di Fisica, Università di Milano)
Description
I discussed some issues that arise when using Machine Learning as an inference tool in the particular context of the determination of parton distributions. Problems I address include: how do we know that the ML model generalizes correctly? Can we detect overlearning? Can we assign an uncertainty to the ML model predictions, and can we validate this assignment?
Auteur principal
Stefano Forte
(Dipartimento di Fisica, Università di Milano)