Göttingen 2025 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
T: Fachverband Teilchenphysik
T 68: Higgs Physics VIII (CP)
T 68.8: Talk
Thursday, April 3, 2025, 18:00–18:15, ZHG105
Symbolic Regression for Higgs CP analyses — •Marco Menen1,2, Henning Bahl3, Elina Fuchs1,2, and Tilman Plehn3 — 1Institut für Theoretische Physik, Universität Hannover, Germany — 2Physikalisch-Technische Bundesanstalt Braunschweig, Germany — 3Institut für Theoretische Physik, Universität Heidelberg, Germany
Additional sources of CP violation beyond those in the Standard Model are needed to produce a sufficient baryon asymmetry of the Universe during baryogenesis. The Higgs sector is an intriguing candidate for such sources and could provide CP violation in the Higgs couplings to fermions and gauge bosons. Recently, much work has been put into optimizing probes of CP violation with machine learning techniques. While such analysis usually outperform analyses of individual observables, the techniques used can be potentially hard to interpret accurately. We demonstrate how different approaches of Symbolic Regression can be used to obtain analytical formula. We then apply our approaches to various steps of a CP analysis, such as the signal-background classification, the classification of the CP state, or the reconstruction of a parton-level observable.
Keywords: Symbolic Regression; Machine Learning; Higgs boson; CP violation