Erlangen 2026 – wissenschaftliches Programm

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

T: Fachverband Teilchenphysik

T 32: Higgs Physics IV

T 32.9: Vortrag

Dienstag, 17. März 2026, 18:15–18:30, KH 01.019

Machine Learning to search for the CP violation of the Higgs boson in H→ττ decays at ATLAS — •Laney Klipphahn, Philip Bechtle, Klaus Desch, Christian Grefe, and Timo Saala — University of Bonn

Reconstruction at ATLAS has been evolving throughout the past decades. Especially machine learning (ML) techniques have been implemented to support and replace individual steps in the reconstruction chain. With these powerful methods, the question arises, to what extend Machine learning models can learn and reproduce the complex calculations that map detector signals to high-level physics observables?

A particularly interesting challenge for ML is the study of CP violation of the Higgs boson in H→ττ decays. The CP property of the Higgs is closely linked to one of the unsolved questions in physics, namely the matter and anti-matter asymmetry observed in the universe today. While the pure CP-odd state has been excluded by measurements, the CP-even or CP-mixed property of the Higgs has yet to be confirmed experimentally. In the Higgs-Yukawa coupling the CP-violation is accessible from the angular correlations of the decay products, which depend on the decay mode of the τ leptons in the H→ττ decay. The CP-property is then obtained by a non-trivial combination of multiple observables in the final state. In this talk I will present the studies on ML for the CP violation of the Higgs boson to assess whether ML models can reach similar or better accuray than traditional reconstruction algorithms.

Keywords: Machine Learning; Higgs CP; ATLAS