Dortmund 2021 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
T: Fachverband Teilchenphysik
T 71: Data analysis, Information technology III
T 71.6: Talk
Wednesday, March 17, 2021, 17:15–17:30, Tu
Performance Studies of the Integration of a Deep-Impact-Parameter-Setsbased Tagger for the ATLAS Experiment b-Tagging Algorithm — •Alexander Froch, Manuel Guth, and Andrea Knue — Albert-Ludwigs Universität Freiburg, Experimentelle Teilchenphysik AG Herten
The identification of the origin of a jet produced in a high-energy collision is
an important task and is crucial for most analyses performed at the ATLAS
experiment. Different multivariate techniques are used and combined to determine
the jet origin. One of these techniques is the Deep Impact Parameter Sets (DIPS)
tagger.
The DIPS tagger is a deep neural network based on the Deep Sets architecture. It
uses track information of the particles inside the clustered jets for
classification. It is part of a new tagging algorithm currently developed in
ATLAS. The algorithm itself can distinguish between different jet origins, like
light, charm or bottom jets. A good performance was already observed for a
training using tt events. For further improvement in the high pT
region, jets from Z′ decays are included in the training.
The performance of this training will be shown along with the impact of the
hyper-parameter optimization studies.