Associated production of the Higgs boson with a W boson in proton-proton collisions: an explorative analysis of the three-leptons final state with the ATLAS experiment

This work concern the study of the properties of the recently discovered Higgs boson in proton-proton collisions at the Large Hadron Collider (LHC). The Higgs boson is the particle that, according to the Standard Model (SM) theory, is responsible for the mass of all the known elementary particles (fermions and bosons). Its existence was postulated during the '60s in the so-called Brout- Englert-Higgs mechanism (BEH), but to have the experimental evidence of this particle scientists had to wait for the first results of the ATLAS and CMS Collaborations at the LHC (2012). Studying the Higgs boson production and decay modes at LHC is one of the purposes of the ATLAS Collaboration. This thesis, in particular, focuses on the study of the Higgs boson production in association with a W boson in proton-proton collision at a centre-of-mass energy of 8 TeV. The channel studied, with the Higgs boson further decaying in two W bosons, allows to direct probe the Higgs boson coupling exclusively with W bosons. The analysis here reported is an explorative study of the three leptons final state in the WH channel, one of the leptons being a hadronically decaying tau. In the first part of the thesis, after a brief description of the SM theory and the BEH mechanism, the ATLAS detector is described and the techniques for the lepton, jet and event identification and reconstruction are presented. The second part of the thesis is about the data analysis done in the study of the WH$\rightarrow$WWW$\rightarrow l \nu l \nu \tau \nu_{\tau}$ (l = e = $\mu$) process. A detailed description of the event selection using Monte Carlo (MC) samples is reported, together with the procedure used to normalize these MC samples to reproduce what observed in data. The cut-based approach was used in the first stage of the analysis, which consists in the application, in sequence, of several selections to the events sample to reduce the background contamination to the signal event topology. Then, due to the low sensitivity of the analysis, a multivariate analysis (MVA) was adopted aimed at reducing the background processes that mimic the signal topology while keeping as much signal events as possible. The last part of the thesis reports the results of the analysis and the comparison with the SM predictions. A binned maximum likelihood fit was used to extract the signal yield as observed in data, and to compare the results obtained with the two analysis approaches with the SM expectations for a Higgs boson of mass m$_H$ = 125 GeV. It is shown that the usage of the MVA approach improved the analysis sensitivity of $\approx$40%. The small statistics, together with the low signal over background ratio of the measurement, do not allow neither to measure nor to exclude the WH process from the SM theory; the upper limit obtained on the WH cross section at 95% CL is 14.5 (9.3 expected). To complete the work prospects for this analysis in ATLAS Run2 data taking period are given.