Bc physics with LHC Run-3 data at the CMS experiment

The work described in this thesis deals with a detailed study focusing on a performant extraction of the signal of the Bc meson in its decay to J/ψ π that is the most suitable decay mode to reconstruct Bc mesons at the Compact Muon Solenoid (CMS) experiment, exploiting the Run-3 data collected in the years 2022-23 at the CERN Large Hadron Collider (LHC). Specifically, large part of the study focuses on the performance comparison between a standard cut-based selection of the Bc signal and one based on Machine Learning (ML), showing the advantage of using XGBoost, namely a parallel Gradient Boosted Decision Trees based library, against a "classical" (non AI-powered) approach. It is also investigated and developed the optimization of the XGBoost hyperparameters exploiting a novel software tool called Optuna.