Classifications of quark/gluon and quenched jets using machine learning technique for high energy heavy ion collisions

In Quantum Chromodynamics, quarks and gluons carry different color charges which lead to different properties of jets. In particular, it is predicted that the energy loss due to the interaction between parton and Quark-Gluon Plasma (QGP) is different for quarks from that for gluons. In order to validate this prediction, it is necessary to distinguish accurately jets originally from quarks (quark jets) from those from gluons (gluon jets). However, the differences between these jets are subtle, making it difficult to distinguish them based on current approaches. Therefore, a machine learning (ML) approach utilizing multivariate analysis is applied. In this study, JETSCAPE[1] framework including parton energy loss is used to study the validity of ML for the jet identification in proton-proton and heavy ion collisions. In this talk, the current status of quenched jet classification for quarks and gluons will be reported.

[1] JETSCAPE collaboration Phys. Rev. C 96 (2017) 024909