Search for the Higgs Boson decaying to tau leptons at ATLAS using multi-variate analysis techniques
Access status:
Open Access
Type
ThesisThesis type
Doctor of PhilosophyAuthor/s
Scarcella, Mark JosephAbstract
This thesis presents three differing approaches to the search for the Standard Model Higgs boson decaying to tau leptons using ps = 7 TeV protonproton collision data from the ATLAS experiment at the LHC. Multi-variate analysis techniques involving boosted decision trees are used ...
See moreThis thesis presents three differing approaches to the search for the Standard Model Higgs boson decaying to tau leptons using ps = 7 TeV protonproton collision data from the ATLAS experiment at the LHC. Multi-variate analysis techniques involving boosted decision trees are used to extend an existing cut-based analysis procedure. The expected 95% confidence level upper limit on the observed cross-section is compared between the analyses. The upper limit at a Higgs mass of mH = 125 GeV is improved from 2:9+4:3 2:1 to 2:3+3:3 1:7 times the Standard Model prediction, after implementing multivariate techniques. No significant excess is seen in data for any analysis strategy. The most sensitive measurement of the signal strength normalised to the Standard Model prediction was observed to be ˆ m = 1:6 1:1, corresponding to 1:4s upward fluctuation of the background-only model to match the data.
See less
See moreThis thesis presents three differing approaches to the search for the Standard Model Higgs boson decaying to tau leptons using ps = 7 TeV protonproton collision data from the ATLAS experiment at the LHC. Multi-variate analysis techniques involving boosted decision trees are used to extend an existing cut-based analysis procedure. The expected 95% confidence level upper limit on the observed cross-section is compared between the analyses. The upper limit at a Higgs mass of mH = 125 GeV is improved from 2:9+4:3 2:1 to 2:3+3:3 1:7 times the Standard Model prediction, after implementing multivariate techniques. No significant excess is seen in data for any analysis strategy. The most sensitive measurement of the signal strength normalised to the Standard Model prediction was observed to be ˆ m = 1:6 1:1, corresponding to 1:4s upward fluctuation of the background-only model to match the data.
See less
Date
2015-12-16Faculty/School
Faculty of Science, School of PhysicsAwarding institution
The University of SydneyShare