Associated Top-Quark-Pair and b-Jet Production in the Dilepton Channel at $\sqrt{s}$ = 8 TeV as Test of QCD and Background to tt+Higgs Production

Two technical and two physics analyses are reported in this thesis, which were√carried out using data recorded by CMS at s = 8 TeV. The first technical contribution is related to the improvement of the performance of the CMS tracking detector,which provides measurements of trajectories of charged particles, and is relevant fornearly every physics analysis at CMS. A new method was studied for Lorentz-angleestimation and correction for the lost signal due to the short readout time betweensubsequent proton-bunch crossings. Such parameters are determined simultaneouslywith the tracker geometry, which allows to include correlations between different effects and to even absorb unknown or mismodelled ones. The obtained results showan improvement of hit-position resolution, which is more stable in time comparedto standalone measurements. Possible ways of further improvement of the methodare also proposed.In the second technical contribution, a standalone software package was developed for the identification of the origin of jets initiated by bottom quarks (b jets)in Monte Carlo (MC) simulations. It allows to differentiate at generator level between b jets originating from decays of e.g. a top-quark, or a Higgs boson, or a Zboson, based on the actual particle relations stored in the simulated event. Thisfunctionality is crucial for the presented physics analyses, and it did not exist inCMS before. Its generic implementation provides equally high performance with allparton-shower generators that are used at CMS, and was chosen as a recommendedway of heavy-flavour-jet definition in future CMS analyses.As one of the primary results of this thesis, the cross sections of top-quark-pairproduction in association with at least one (ttb(b)) or at least two (ttbb) b jets withpT > 20 GeV were measured differentially as functions of transverse momentum(pT ) and absolute pseudorapidity (|η|) of the first or the second additional b jetrespectively. The ttbb cross section was also measured as a function of the angulardistance (∆Rbb ) and of the invariant mass (mbb ) of the two additional b jets. Thelatter provides substantial sensitivity to the ttH process, to which ttbb is an almostirreducible background, therefore the data is blinded in the ttH-enhanced invariantmass region. The dileptonic final state of the tt process is considered to minimisethe probability of possible jet misassignments. The measured ttb(b) (ttbb) crosssection is about 1.9 (1.3) times higher than predicted by MC simulations, which isin agreement with the previous inclusive ttbb cross-section measurement by CMS.The shapes of the measured cross sections are well modelled by MC predictions,except for the ∆Rbb spectrum, which shows a tendency towards smaller simulatedseparation between additional b jets than measured in the data.The measured ttb(b) and ttbb cross sections were used to estimate the sensitivityto the ttH(bb) process based on MC simulations, without systematic-uncertaintiestreatment. Several approaches of candidate jet-pair selection were studied, whichaim at the extraction of the ttH signal from the mbb distribution, which is a wellunderstood physical quantity. The highest signal-over-background ratio reached is1/9, at which the ttH-signal significance is 0.18σ for the available 19.7 fb−1 of data.√A rough estimate of the sensitivity with future data recorded at s = 14 TeV showedthat about 130 fb−1 of data should be sufficient to see a mass peak from H → bbdecays with a significance of 2σ.