Triple Higgs production at LHC using CMS data, decaying into the "bb̅bb̅τ⁺τ⁻" final state

Details

Supervisors

Bethani, Agni

Faculty

Faculté des sciences

Degree label

Master [120] en sciences physiques, à finalité approfondie

Abstract

enIn this thesis, we studied and tested an analysis strategy for the bb̅bb̅τ⁺τ⁻ final state, with the objective of training and using machine learning classifiers to discriminate triple Higgs events from other backgrounds. We first presented the basic variables collected by CMS and simulated by the Monte Carlo algorithm. We then identified and computed additional useful variables, to further improve the discrimination quality of the classifiers. These classifiers were then trained and tested against various backgrounds, such as top-quark pair, vector boson fusion double Higgs and gluon-gluon fusion double Higgs. The classifiers are subsequently used on these backgrounds, which allows us to compute their signal to background ratios for various cuts.