BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:UW-Madison-Physics-Events
BEGIN:VEVENT
SEQUENCE:1
UID:UW-Physics-Event-6328
DTSTART:20210310T170000Z
DTEND:20210310T181500Z
DTSTAMP:20260415T023553Z
LAST-MODIFIED:20210307T162815Z
LOCATION:Online Seminar: Please sign up for our mailing list at www.ph
 ysicsmeetsml.org for zoom link
SUMMARY:Generative and Invertible Networks for the LHC\, Physics ∩ M
 L Seminar\, Tilman Plehn\, Heidelberg University
DESCRIPTION:LHC physics is a unique field in the sense that we compare
  vast and highly complex data sets with precise first-principles predi
 ctions. These predictions usually rely on Monte Carlo simulations. I w
 ill show how generative neural networks can supplement these simulatio
 ns and discuss conceptional advantages of this method. I will then exp
 lain how generative networks can invert event simulations. Flow-based 
 invertible networks allow us to invert or unfold individual detector s
 imulations of QCD parton showers in a mathemacially consistent manner.
  That means that they predict calibrated probability distributions in 
 parton-level phase space for individual observed events. Finally\, I w
 ill illustrate how the same networks can infer the structure of QCD sp
 littings forming jets.
URL:https://www.physics.wisc.edu/events/?id=6328
END:VEVENT
END:VCALENDAR