Choose timezone
Your profile timezone:
Description
Reliably estimating hadronization effects is an important aspect of high-precision collider program. While infrared and collinear safe observables are in principle not sensitive to hadronization, this is no longer the case in practice, due to the presence of detector effects and experimental cuts. Properly modeling hadronization thus becomes an important systematic in many instances, for which we need to resort to modeling. In this talk I will review the progress made by the MLhad collaboration on a data-driven approach to describing hadronization using machine learning. For this one wants to move away from infrared and collinear safety and find observables, that are instead most sensitive to hadronization (thus also “most IRC unsafe”) as new inputs to the ML training. Examples include charge and hadron multiplicities, which have been measured in the past, but also multiplicity flow correlators, and others, that have not yet.
