Measuring the b-tagging efficiency in ATLAS data at $\sqrt{s} = 8$ TeV and $\sqrt{s} =$ 13 TeV using the $p_{\mathrm{T}}^{\mathrm{rel}}$-method

This thesis presents the first measurement of the $b$-tagging efficiency $\varepsilon_b$ with the $p_{\mathrm{T}}^{\mathrm{rel}}$-method under Run 2 conditions at the ATLAS experiment. The measurement has been performed for semi-leptonically decaying $b$-hadrons with a muon in the final state. This has been achieved by selecting events with a muon associated to a jet by a cone based matching algorithm. The $p_{\mathrm{T}}^{\mathrm{rel}}$-variable is sensitive to the mass of the weakly decaying particle and is therefore sensitive to the jet-flavor. Exploiting this dependence, the flavor fractions are determined with a log-likelihood template fit to the $p_{\mathrm{T}}^{\mathrm{rel}}$-distribution in the tagged and untagged jet samples determined by the algorithm and operating point under study. The dataset used for this study corresponds to an integrated luminosity of 11.6 fb$^{-1}$ of proton-proton-collisions taken with the ATLAS detector at $\sqrt{s} = 13$ TeV center of mass energy. In addition, $\varepsilon_b$ has been measured for algorithms on trigger level using the $p_{\mathrm{T}}^{\mathrm{rel}}$-method. This has been conducted in a set of proton-proton-collisions recorded at the ATLAS experiment at $\sqrt{s} = 8$ TeV center of mass energy. Here, a conditional calibration depending on off-line tagger decisions has been utilized.