Karlsruhe 2024 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 94: Trigger+DAQ 3

T 94.2: Talk

Thursday, March 7, 2024, 16:15–16:30, Geb. 30.23: 3/1

Development of machine-learning based topological algorithms for the CMS level-1 trigger — Johannes Haller, Gregor Kasieczka, Karla Kleinbölting, •Finn Labe, Artur Lobanov, Matthias Schröder, and Shahin Sepanlou — Institut für Experimentalphysik, Universität Hamburg

Using a HH process as an example, the possibility of using machine learning to construct trigger selections using full event topologies is studied. Targeting the CMS level-1 trigger, it is shown that simple neural networks can provide increased sensitivity at low rate costs and that these neural networks can be deployed in the FPGA-based electronics of the trigger system.

Keywords: Machine Learning; L1 trigger; Topological trigger; Integration; Di-Higgs