Speaker
José Manuel Clavijo Columbié
(DESY)
Description
The ATLAS ttH(bb) analysis is limited by the background modelling uncertainties, that result in a bias of the classifier towards the Monte Carlo generator used for training. We apply adversarial domain adaptation to train a more generic classifier. We employ a neural network that simultaneously classifies the signal versus background events, while minimizing the difference of the classifier response to two alternative background MC models.
Author
José Manuel Clavijo Columbié
(DESY)
