Studying a denition for a boosted W/Z/H jet tagger at the FCChh, employing modern Machine Learning algorithms and customised features (beyond the usual substructure variables)

A jet is a spray of particles, usually produced by the hadronization of a quark or gluon in a particle physics or heavy ion experiment. Reconstructed particles are clustered into jets using one of the available jet clustering algorithms (kT, anti-kT etc.), which adopt dierent metrics to decide if two given particles belong to the same jet or not. Jets can also originate from the decay of high-momenta heavy particles, such as boosted vector boson. When these particles decay to quarks, the overlap of the hadronization products of each quark result into a single massive jet, dierent than the ordinary jets from quarks and gluons. These special jets can be identied using substructure algorithms. In this study, we consider the performances of a commonly used substructure variable, N-subjettiness, with two variants of an alternative approach, based on the momentum ow around the jet axis. I focused on high-energy collision in a hypothetical future circular collider (FCC) colliding protons at a center-of-mass energy 100 TeV.