guest ::
| Home > Publications database > Studien zur Bestimmung der Di-Higgs-Masse mitHilfe neuronaler Netze |
| Home > Publications database > Studien zur Bestimmung der Di-Higgs-Masse mitHilfe neuronaler Netze |
| Master Thesis | PUBDB-2025-04942 |
; ;
2025
Abstract: (di-Higgs) events using simulated data from the ATLAS detector at the Large Hadron Collider ata center-of-mass energy of √s = 13TeV and an integrated luminosity of 140 fb−1. In the studiedchannel, one Higgs boson decays into a bottom quark-antiquark pair (b¯b), the other into two photons(γγ). Direct measurement of Higgs self-coupling through di-Higgs processes serves as a consistencytest for the Higgs potential in the Standard Model and may indicate physics beyond it. Sensitivityto the coupling modifier for Higgs self-coupling depends on the di-Higgs mass spectrum resolution,which is limited by the energy resolution of bottom-quark jets and missing energy from undetectableneutrinos in the decays. The first method uses missing transverse energy to correct the energiesof bottom-quark jets. The second employs a neural network to regress the di-Higgs mass fromreconstructed observables. For events where the correction is applicable, the jet energy correctionimproves the b¯b mass distribution’s median value by about 9.6 GeV compared to the calibration inthe current ATLAS analysis. The neural network reduces the standard deviation of the relativedeviation from the true di-Higgs mass by approximately 23% and the interquartile range by about3% compared to the kinematic fit used in the current ATLAS analysis.II
|
The record appears in these collections: |
PDF
PDF (PDFA)