Searches for resonant double Higgs production in the bbττ final state at the CMS experiment

This thesis presents a search for resonant Higgs bosons pair (HH) production via the Gluon-gluon Fusion (ggF) mechanism, predicted in different Beyond Standard Model (BSM) scenarios to be observable with the current Large Hadron Collider (LHC) dataset. The search focuses on the HH → b¯bτ+τ− final state, which offers a good balance between branching fraction and experimental feasibility but presents challenges due to neutrinos in τ decays and background discrimination. Advanced techniques are employed for τ reconstruction and identification, along with sophisticated modeling of the relevant backgrounds. The analysis, performed on proton-proton (pp) collision data collected by the Compact Muon Solenoid (CMS) experiment during Run 2 (2016–2018) with an integrated luminosity of 137.1 fb−1, employs advanced techniques for τ reconstruction and background modeling. The τ+τ− pair is studied through three decay channels: τhτh (fully hadronic) and τhτℓ, where τh represents a tau lepton decaying into hadrons plus a ντ , and τℓ corresponds to the leptonic decay of the tau into either an electron (ℓ = e) or a muon (ℓ = µ). A key aspect of the analysis is the use of machine- learning-based identification algorithms, including DeepTau for τ identification. The primary results are obtained using DeepTau2017v2p1, the stable version recommended for Run 2 analyses. Additionally, a study with DeepTau2018v2p5, an improved and re-trained version optimized for future runs, demonstrates better data/MC agreement, confirming its suitability for upcoming analyses. As part of the CMS internal review process, the analysis remains blinded to avoid bias. The 95% Confidence Level upper limits on the HH production cross section are set as a function of resonance mass and spin, interpreted within the Warped Extra Dimensions theory (WED) model for masses from 250 to 3000 GeV under spin-0 and spin-2 hypotheses.