Search for Higgs-Boson pair-production in the bbℓℓ + ETmiss final state with the ATLAS detector at sqrt(s) = 13 TeV : = Search for Higgs-Boson Pair-Production in the bbℓℓ + ET miss final state with the ATLAS detector at [square root]s = 13 TeV

Abstract: This thesis presents the measurement of the Higgs-boson pair-production cross-section in the ggF and VBF production modes σ(ggF+VBF,HH) and the corresponding signal strength µ(HH) = σ(ggF+VBF,HH)/σ(ggF+VBF,HH,SM) using the bbℓℓ+ETmiss final state. Furthermore, a constraint on the Higgs-boson self-coupling modifier κ(λ) = λ/λ(SM) is extracted, where λ denotes the coupling strength of the trilinear Higgs-boson self-interaction. The dataset analyzed in this thesis corresponds to proton-proton collisions at a center-of-mass energy of 13 TeV that have been produced by the Large Hadron Collider and have been collected with the ATLAS detector during Run-2 (2015-2018). The integrated luminosity of the dataset is 139 1/fb.

The bbℓℓ+ETmiss final state includes all decay chains that lead to two b-tagged jets, two light leptons, and some amount of missing transverse energy. It can be produced via the HH→bbWW→bbℓℓ+2ν, HH→bbττ→bbℓℓ+4ν, and HH→bbZZ→bbℓℓ+2ν/4ν/2q decay chains. The bbℓℓ+ETmiss final state profits from the large branching ratio of the H→bb decay and the good efficiency of lepton triggers.

Dominant background contributions arise from single top-quark production and top-quark pair production, as well as Z-boson production in association with heavy flavor jets. The normalization of these backgrounds is determined from data from three control regions. Events that contain misidentified or non-prompt leptons are estimated with a data-driven method. Minor background contributions are taken from simulation.

Two signal regions (SR 1 and SR 2) that are enriched in signal events have been defined, targeting different parts of the signal process. A multi-class neural network has been trained for each signal region to further separate signal from background. The separation performance of the neural networks has been optimized by tuning the hyperparameters via the method of Bayesian hyperparameter optimization. Furthermore, the set of input features has been optimized by ranking the input features by their importance, which is given by the permutation feature importance algorithm. The distribution of the signal output node of the neural network is used as the discriminant for the statistical analysis.

The cross-section of Higgs-boson pair-production is measured to be σ(ggF+VBF,HH)=(-50+/-160)fb with a corresponding signal strength of µ(HH)=-1.52 +5.02/-5.15. Upper limits at 95% confidence level are set at 287fb for σ(ggF+VBF,HH) and 9.7 for µ(HH), with an expectation of 304fb and 10.3, respectively. A constraint is set on κ(λ) with an observed interval of -6.32<κ(λ)<14.21 and an expectation of -6.30<κ(λ)<13.94

Angaben zum Objekt

• Weitere Titel: Search for Higgs-Boson Pair-Production in the bbℓℓ + ET miss final state with the ATLAS detector at [square root]s ; 13 TeV
• Standort: Deutsche Nationalbibliothek Frankfurt am Main
• Umfang: Online-Ressource
• Sprache: Englisch
• Anmerkungen: Universität Freiburg, Dissertation, 2023

Klassifikation und Themen

• Schlagwort: ATLAS ; Higgs-Teilchen ; Elementarteilchenphysik ; Elektroschwache Wechselwirkung ; Standardmodell ; t-Quark ; Elementarteilchenphysik ; Higgs-Teilchen ; LHC ; ATLAS

Beteiligte, Orts- und Zeitangaben

• Ereignis: Veröffentlichung
• (wo): Freiburg
• (wer): Universität
• (wann): 2023
• Urheber: Rottler, Benjamin
• Beteiligte Personen und Organisationen: Schumacher, Markus ; Schumacher, Markus ; Fischer, Horst ; Albert-Ludwigs-Universität Freiburg. Fakultät für Mathematik und Physik ; Albert-Ludwigs-Universität Freiburg. Physikalisches Institut ; Experimentelle Teilchenphysik, Abt. Prof. Schumacher ; Albert-Ludwigs-Universität Freiburg. Fakultät für Mathematik und Physik

Weitere Informationen

• DOI: 10.6094/UNIFR/238604
• URN: urn:nbn:de:bsz:25-freidok-2386046
• Rechteinformation: Open Access; Der Zugriff auf das Objekt ist unbeschränkt möglich.
• Letzte Aktualisierung: 25.03.2025, 13:45 MEZ