Unit:
Κατεύθυνση Πυρηνική Φυσική και Φυσική Στοιχειωδών ΣωματιδίωνLibrary of the School of Science
Author:
Katris Panagiotis
Supervisors info:
Κωνσταντίνος Βελλίδης, Aναπληρωτής Καθηγητής, Τμήμα Φυσικής, ΕΚΠΑ
Original Title:
Implementation of a lepton identification algorithm using machine learning for measurements of the tt+X associated production process in the CMS experiment at the LHC
Translated title:
Implementation of a lepton identification algorithm using machine learning for measurements of the tt+X associated production process in the CMS experiment at the LHC
Summary:
The goal of this thesis is the development of techniques for the identification and
selection of leptons originating from top quark decays via a subsequent leptonic W bo-
son decay. In the associated production of top quarks with vector bosons, prompt lep-
tons can also emerge from the associated vector boson decay. Leptons from hadronic
decays are considered as non-prompt and appear to be less isolated than the prompt
ones. These non-prompt leptons are among the most important backgrounds in pro-
cesses that have multilepton final states. This thesis focuses on the implementation
of a machine learning algorithm for lepton discrimination in tt+X processes at CMS
with emphasis to the t t̄H(cc̄), aiming to reduce the non-prompt lepton background
and retaining at the same time maximum prompt lepton selection efficiency.
Main subject category:
Science
Keywords:
leptons, machine learning, Higgs boson, identitification, background, prompt
