Jet Tagging with HGCal

Zheng-Gang Chen^1*, Kai-Feng Chen¹

¹Department of Physics, National Taiwan University, Taipei, Taiwan

* Presenter:Zheng-Gang Chen, email:a0910555246@gmail.com

Compact Muon Solenoid (CMS) experiment plans to update its end-cap detector with a powerful design, High Granularity Calorimeter (HGCal) in 2026. This novel design not only increases a considerable resolution in the transverse direction, but also provide longitudinal information of particles passing it. And this can benefits many fields of particle physics, jet tagging analysis is also included. Traditionally, we have information that is based on 2D detectors, and this also give a restriction on classification task in quark and gluon separation, which is said to be hard to tell the difference based on detector view. But not only with additional dimensional information and resolution improvement in transverse direction, the detectors also provide us time resolution, so let's said we have 5D information, which is 3D spatial, 1D Energy and 1D Time. With this great design, this talk will present how we get improvement on the performance of q-g tagging.

Keywords: HGCal, Quark and Gluon, Convolutional Neural Network, Machine Learning, CMS