Aachen 2019 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 53: Suche nach Neuen Teilchen III

T 53.6: Vortrag

Mittwoch, 27. März 2019, 17:15–17:30, H07

Reconstruction techniques for displaced particle decays in the CMS calorimeter — Gregor Kasieczka and •Zhiyuan He — Institut für Experimentalphysik, Universität Hamburg, Deutschland

We present how deep neural networks (DNN) can be exploited to improve searches for beyond Standard Model (SM) physics, peformed with the CMS detector. One of the SM issues, the hierarchy problem, can be addressed by introducing a so-called "dark sector", that interacts with SM particles via the Higgs boson. The Higgs boson is expected to decay into a pair of new scalar long lived particles (LLPs), that in turn decay predominantly into SM b quark pairs, expected to be "displaced" with regards to the interaction point. If the new scalar particles have decay lengths of approximately the radius of the CMS calorimeter radius, high-level jet variables, as well as low-level quantities calculated from jet components, can be used to discriminate the b-jet decay products of the LLPs from SM b-jets. These attributes, extracted from Monte Carlo simulations of the new exotic process and of SM processes, are used to train a DNN algorithm. We focus on optimizing the search strategy for masses of the LLP ranging from 15 to 60 GeV, and decay lengths from 100 mm up to several meters.