Probing Quark Hadronization with B mesons at the LHC

In nature, quarks cannot exist by themselves. When produced, like for instance in high energy collisions, quarks join together with other quarks to form hadrons. This process, referred to as hadronization, is driven by the strong force (QCD) and is yet not fully understood. Measurements of B mesons at colliders offer unique probes of the hadronization process by which single quarks form color-neutral hadrons. In this work, $B^+$ and $B^0_s$ meson signals are studied and identified in the LHC collisions, employing state-of-the-art machine learning methods. The B mesons are reconstructed through the decay channels "$B^{+} \rightarrow J/\psi ~K^{+} \rightarrow \mu^{+}\mu^{-}K^{+}$" and $B^{0}_{s} \rightarrow J/\psi ~\phi \rightarrow \mu^{+}\mu^{-}K^{+}K^{-}$, in proton-proton (pp) collision data at $\sqrt{s}=5.02$ TeV collected in 2017 by the Compact Muon Solenoid experiment (CMS) at the LHC. This thesis goes over all the developed work that encompasses a full data analysis. This includes the event selection, achieved here by utilizing the XGBoost package for the creation of a boosted decision tree model, parameter fitting of the obtained invariant mass distributions, and detector and selection efficiency determination, in which we apply a novel method that minimizes the reliance on Monte Carlo simulation kinematics and removes background contamination in the signal region. We also present performance comparisons between BDT and neural network (NN) models, along with studies that aim at probing the more challenging low $p_T$ region. With these ingredients, we measure the B mesons' differential production cross section as function of signal kinematic variables (transverse momentum and rapidity) and environment variables (charged-particle multiplicity), obtaining results consistent with FONLL theoretical predictions and previous results. We calculate the cross section ratios between the two B mesons, investigating dependencies of the fragmentation fraction ratio, $f_s/f_u$, on these variables. The results provide hints of $B^0_s$ production enhancement relative to $B^+$ at high charged-particle multiplicities, as would be expected when considering quark coalescence as a complementary hadronization mechanism.