Measurement of the branching ratios for the Standard Model Higgs decays into muon pairs and into Z boson pairs at a 1 . 4 TeV CLIC

The measurement of the Higgs production cross-section times the branching ratios for its decays into μ+μ and ZZ⇤ pairs at a 1.4 TeV CLIC collider is investigated in this paper. The Standard Model Higgs boson with a mass of 126 GeV is dominantly produced via WW fusion in e+e collisions at 1.4 TeV centre-of-mass energy. Analyses for both decay channels are based on a full simulation of the CLIC_ILD detector. All relevant physics and beam-induced background processes are taken into account. An integrated luminosity of 1.5 ab 1 and unpolarised beams are assumed. For the H ! ZZ⇤ decay, the purely hadronic final state (ZZ⇤ ! qq̄qq̄) is considered as well as ZZ⇤ decays into two jets and two leptons (ZZ⇤ ! qq̄l+l ). It is shown that the branching ratio for the Higgs decay into a muon pair times the Higgs production cross-section can be measured with 38% statistical uncertainty. It is also shown that the statistical uncertainty of the Higgs branching fraction for decay into a Z boson pair times the Higgs production cross-section can be measured with a precision of 18.3% and 5.6% for the hadronic and semi-leptonic ZZ⇤ decays, respectively. Talk presented at the 9th International Physics Conference of the Balkan Physical Union (BPU9), Istanbul, Turkey, 24–27 August 2015. 1)gordanamd@vinca.rs Measurement Of The Branching Ratios For The Standard Model Higgs Decays Into Muon Pairs And Into Z Boson Pairs At A 1.4 TeV CLIC Gordana Milutinović-Dumbelović1, a), Ivanka Božović-Jelisavčić1) , Christian Grefe2,3), Goran Kačarević1), Strahinja Lukić1), Mila Pandurović1), Philipp Roloff2), Ivan Smiljanić1) On behalf of the CLICdp collaboration 1Vinca Institute of Nuclear Sciences, University of Belgrade, Mihajla Petrovića Alasa 12-14, 11001 Belgrade, Serbia 2CERN, CH-1211 Geneva 23, Switzerland 3Universität Bonn, D-53012 Bonn, Germany a)Corresponding author: gordanamd@vinca.rs Abstract. The measurement of the Higgs production cross-section times the branching ratios for its decays into μ+μand ZZ∗ pairs at a 1.4 TeV CLIC collider is investigated in this paper. The Standard Model Higgs boson with a mass of 126 GeV is dominantly produced via WW fusion in e+ecollisions at 1.4 TeV centre-of-mass energy. Analyses of both decay channels are based on a full simulation of the CLIC_ILD detector. All relevant physics and beaminduced background processes are taken into account. An integrated luminosity of 1.5 ab-1 and unpolarised beams are assumed. For the H→ZZ* decay, the purely hadronic final state (ZZ∗ → q?̅?q?̅?) is considered as well as ZZ∗ decays into two jets and two leptons (ZZ∗ → qql+l−). It is shown that the branching ratio for the Higgs decay into a muon pair times the Higgs production cross-section can be measured with 38% statistical uncertainty. It is also shown that the statistical uncertainty of the Higgs branching ratio for decay into a Z boson pair times the Higgs production crosssection can be measured with a precision of 18.3% and 5.6% for the hadronic and semi-leptonic ZZ∗ decays, respectively.


INTRODUCTION
The Compact Linear Collider (CLIC) is an option for a future multi-TeV linear electron-positron collider.One of the main aims of CLIC would be the high-precision measurement of the Higgs boson properties [1,2].In e + e -collisions at the centre-of-mass energy √ =1. 4 TeV the SM-like Higgs boson with a mass of 126 GeV is dominantly produced via WW fusion with ~370000 expected events in 1.5 ab −1 of data.This would lead to the measurements of the couplings of the Higgs boson to the electroweak gauge bosons at the percent level as well as give access to rare processes such as  →  +  − .

Event Samples
The cross-section for Higgs production in WW fusion at 1.4 TeV is 244 fb without beam polarization.The  →  +  − signal statistics are expected to be small (of the order of a few tens of events) because of the small branching ratio for this particular decay.In Table 1, the full list of physics backgrounds is given.

Preselection And Multivariate Analysis
The preselection in the analysis requires the reconstruction of a muon and an antimuon in an event, a di-muon invariant mass,   , in the range (105-145) GeV, the absence of a high-energy electron (E>200 GeV) and a polar angle above 30 mrad.For the final selection, multivariate analysis (MVA) techniques are used based on the boosted decision tree (BDT) classifier.The following 6 discriminating observables are used for the classification of events: visible energy of the event excluding the energy of the di-muon system (Evis), transverse momentum of the di-muon system (pT(µμ)), scalar sum of the transverse momenta of the two selected muons (pT(µ1)+ pT(µ2)), boost of the di-muon system (β(µμ)), polar angle of the di-muon system (θ(µμ)), cosine of the helicity angle (cosθ*).The cut on the classifier output was optimised to minimise the statistical uncertainty of the measurement.The MVA selection efficiency for the signal is 32%.The overall signal efficiency including reconstruction, preselection, losses due to coincident tagging of Bhabha particles and the MVA is 24%, resulting in an expected number of 19 signal events.Distributions of the di-muon invariant mass before and after the MVA selection are shown in Figure 1.

Di-Muon Invariant Mass Fit
In order to determine (    ̅ ) × ( →  +  − ), the number of selected signal events Ns has to be known.The number of signal events is determined by fitting the probability density functions (PDFs) describing the signal and the background distributions of the di-muon invariant mass.In order to estimate the statistical uncertainty of the measurement, 5000 toy Monte Carlo (MC) experiments are performed.The RMS of the distribution of the number of signal events per experiment corresponds to the statistical uncertainty of the measurement of 38% [5].

Event Samples
In this analysis, two final states are studied: the fully hadronic final state,  * →  ̅ ̅, with a branching ratio of 49% resulting in an effective cross section of 3.45 fb and the semi-leptonic final state,  * →  +  − , with a branching ratio of 10% and an effective cross-section of 0.995 fb.
In Table 1, the list of the signal and main background processes is given with the corresponding cross sections.

Preselection And Multivariate Analysis
For the semi-leptonic final state, the first step of the physics object identification is searching for isolated leptons (electrons, muons or taus).Exactly two leptons are required, otherwise the event is rejected.Then, all particles in the event not identified as leptons are clustered by the kt algorithm into two jets.For the hadronic final state, the event is directly clustered by the kt algorithm into four jets.
For both final states, flavour-tagging is performed and a preselection based on kinematic variables is applied.The preselection cuts for the fully hadronic final state require one on-shell Z boson (45 GeV < mZ < 110 GeV), one offshell Z boson (mZ*< 65 GeV), a Higgs invariant mass in the range (90 GeV < mH < 165 GeV), the distance value between the two closest jets (−log y34 < 3.5, −log y23 < 3.0), a visible energy in the range (100 GeV < Evis < 600 GeV), a missing transverse momentum above 80 GeV (p T miss>80 GeV) and b-tag probabilities of less than 0.95 for both jets (P(b) ( jet 1 ) , P(b) ( jet 2 ) <0.95).After the preselection, an MVA event selection based on the BDT classifier is performed to obtain the final results.In both final states, the BDT cut maximising the significance is chosen, giving an overall efficiency of 18% and 30.4%, for the fully-hadronic and semi-leptonic final states, respectively.Figure 3 (a) includes all events that pass the preselection, while Figure 3 (b) shows all events passing the BDT selection for semi-leptonic final state.The statistical uncertainty for (    ̅ ) × ( →  * ) is calculated as: where S denotes the number of selected signal events, and B the number of selected background events.The final result is found to be 18.3% and 5.6% for the fully-hadronic and semi-leptonic final state, respectively.

RESULTS AND CONCLUSIONS
The expected precisions for the measurements of (    ̅ ) × ( →  +  − ), (    ̅ ) × ( →  * →  ̅ ̅) and (    ̅ ) × ( →  * →  ̅ +  − ) are found to be 38%, 18.3% and 5.6%, respectively.They are dominated by the limited signal statistics and the presence of large backgrounds.The obtained results are included in the fit to all possible measurements at CLIC to contribute to the Higgs to Z coupling, Higgs to µ coupling and to the total Higgs width ГH.

FIGURE 1 .
FIGURE 1. Stacked histograms of the di-muon invariant mass distributions with preselecton only (a) and after MVA selection (b).

FIGURE 2 .
FIGURE 2. Fitted distribution of the di-muon invariant mass mμμ, for the sum of the signal and the total background in one toy MC experiment.

FIGURE 3 .
FIGURE 3. The Higgs invariant mass distributions after preselection (a) and after MVA selection (b)for the semi-leptonic final state.

TABLE 1 .
List of considered processes with corresponding cross sections.