Invertible Networks or Partons to Detector and Back Again

2020/06/30 by Marco Bellagente, Anja Butter, Gregor Kasieczka +5 · 1 citation
Physics and Astronomy · #hep-ph

paper · pdf · doi:10.21468/scipostphys.9.5.074

published as SciPost Phys. 9, 074 (2020) · 25 pages, 10 figures

arxiv created 2020/10/01 · arxiv updated 2020/11/18

Abstract

For simulations where the forward and the inverse directions have a physics meaning, invertible neural networks are especially useful. A conditional INN can invert a detector simulation in terms of high-level observables, specifically for ZW production at the LHC. It allows for a per-event statistical interpretation. Next, we allow for a variable number of QCD jets. We unfold detector effects and QCD radiation to a pre-defined hard process, again with a per-event probabilistic interpretation over parton-level phase space.

Cited by

• Unbinned Inference with Correlated Events
• Machine Learning is Good for Physics - and Vice Versa