Estimating deceiving signatures and their role in the observation of the VBF production mode in the Higgs-Boson decay into two W bosons

Date created: 
ATLAS experiment
Higgs boson
Higgs to WW decay
Fake leptons
Fake factor method
Machine learning

The Higgs boson is the most recently discovered fundamental particle and its precise characterization is an important test of the Standard Model. For these measurements, the decay into two W bosons is crucial because of the large branching ratio, the sensitivity to vector-boson couplings and the theoretical implications on WW scattering. The proton-proton collisions at the LHC with a centre-of-mass energy of √s = 13 TeV provide an excellent environment to study the Higgs boson. The data recorded by the ATLAS experiment between 2015 and 2018 is analyzed. With an integrated luminosity of 139/fb, it allows to measure precisely the two dominant production modes, gluon fusion (ggF) and vector-boson fusion (VBF). Only collisions with one electron and one muon in the final state are considered and cuts are applied to select signal-like events. For the VBF process, the selection is further refined with the help of a deep neural network. Then, the purified dataset is analyzed with a statistical model constructed from MC simulation and a data-driven background estimate of misidentified leptons. The cross sections times branching ratio are measured to be 12.4 ± 1.5 pb for the ggF and 0.79 +0.19 −0.16 pb for the VBF processes. For the first time, the VBF process is observed in this decay channel with a significance of 6.6 standard deviations. Also measurements of eleven Simplified Template Cross Sections are performed and found to be consistent with the Standard Model. This precise measurement could only be achieved because the treatment of events with misidentified leptons was improved significantly. These deceiving signatures are estimated with the data-driven Fake Factor Method. An auxiliary measurement is performed to extract fake factors in a phase space with a Z boson recoiling against a jet that is misidentified as lepton. This measurement is presented together with the relevant experimental improvements. As a result of this work, the uncertainties could be reduced to an extent that they do not significantly affect the measurement of the Higgs-boson cross section. Even though the Fake Factor Method is experimentally well-established, a consistent derivation for a phase space with many leptons does not exist. In this thesis, it is shown for the first time that the Fake Factor Method can be derived exactly from the Matrix Method for an arbitrary number of misidentified leptons. In addition, the uncertainty estimate is considered and cases are derived in which the fake estimate does not exhibit Gaussian uncertainties.

Document type: 
This thesis may be printed or downloaded for non-commercial research and scholarly purposes. Copyright remains with the author.
Bernd Stelzer
Science: Department of Physics
Thesis type: 
(Thesis) Ph.D.