Evidence for the Production of the Standard Model Higgs Boson Produced via Vector Boson Fusion in the WW* Channel at the ATLAS Detector

Date created: 
Standard Model Higgs Boson
Vector Boson Fusion Higgs production
Higgs to WW decay
Machine learning
Boosted Decision Tree
ATLAS experiment

In 2012, the ATLAS and CMS experiments at CERN's Large Hadron Collider announced they had each observed a new particle with a mass of about 125 GeV/c^2. Given the available data, the properties of this particle are consistent with the Higgs boson predicted by the Standard Model of particle physics (SM). The Higgs boson, as proposed within the SM, is the simplest manifestation of the Brout-Englert-Higgs mechanism. This discovery was driven by the gluon fusion (ggF) production mode, the dominant Higgs boson production mechanism at the LHC. The SM also predicts that the Higgs boson can be produced by the fusion of two weak vector bosons (VBF). Measuring VBF Higgs boson production is an important test of the SM but it is challenging to measure given its cross section is an order of magnitude smaller than that of ggF. After H->bb, H->WW* is the dominant decay channel for the SM Higgs boson at 125 GeV/c^2 and is therefore a promising channel to measure its properties. In addition, the VBF H->WW* search channel makes it possible to probe the exclusive coupling of the Higgs boson to the weak vector bosons. Precise measurements of these coupling strengths make it possible to constrain new models of physics beyond the SM. Despite its relatively large branching ratio, H->WW*->lnln is a challenging channel to search for the Higgs boson because of the neutrinos in the final state which are not directly detectable by the ATLAS detector. Consequently, it is not possible to fully reconstruct the mass of the WW system. Furthermore, there are several backgrounds that have the same signature in the detector as the signal. Top quark pair production is the largest background in this analysis. A multivariate analysis technique, based on an eight-variable boosted decision tree (BDT), is used to search for VBF H->WW*->lnln in the Run-I data and a subset of the Run-II data. This analysis provides the first evidence for VBF H->WW*->lnln with a significance of 3.2 standard deviations in Run-I and 1.9 standard deviations in Run-II. The measured signal strength relative to the rate predicted by the SM for VBF H->WW*->lnln is 1.3 +/- 0.5 using the Run-I data, and 1.7 +1.1/-0.9 using a fraction of the Run-II data.

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