SMuK 2023 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
T: Fachverband Teilchenphysik
T 60: Theory BMS
T 60.5: Talk
Wednesday, March 22, 2023, 16:50–17:05, HSZ/0201
Constraining BSM scalars with neural networks — Thomas Flacke1, Jeong Han Kim2, •Manuel Kunkel3, Jun Seung Pi2, Werner Porod3, and Leonard Schwarze3 — 1Center for AI and Natural Sciences, KIAS, Seoul, Republic of Korea — 2Department of Physics, Chungbuk National University, Chungbuk, Republic of Korea — 3Institut für Theoretische Physik und Astrophysik, Julius-Maximilians-Universität Würzburg, Germany
We study a simple extension of the Standard Model motivated by composite Higgs models, in which a doubly charged scalar decays to W+ t b, resulting in a 4t-like signature from pair production. We train a neural network to differentiate this BSM signal from the dominant SM backgrounds using jet images and kinematic data. We derive the discovery reach and expected exclusion limit at the LHC. A comparison with recasts of Run-2 analyses shows a significant improvement over cut-based analyses.