Bonn 2020 – wissenschaftliches Programm
Die DPG-Frühjahrstagung in Bonn musste abgesagt werden! Lesen Sie mehr ...
T 14.5: Vortrag
Montag, 30. März 2020, 17:35–17:50, H-HS XV
Demonstrating learned particle decay reconstruction with graph neural networks at Belle II — •Ilias Tsaklidis1, James Kahn2, Tobias Boeckh2, and Pablo Goldenzweig2 for the Belle II collaboration — 1Institut Pluridisciplinaire Hubert CURIEN (IPHC), Strasbourg, France — 2Karlsruher Institut für Technologie, Germany
The clean environment within Belle II, with decay processes originating from an electron-positron pair without the presence of partons, allows for the reconstruction of the entire collision event. This is a unique advantage to the Belle II experiment in that it allows for direct measurements of decay processes involving neutrinos or few detectable particles in the final state. This does, however, require a catch-all reconstruction algorithm which is able to reconstruct those particles not associated with the signal process being investigated. The current Full Event Interpretation algorithm at Belle II requires the reconstructed sub-decay processes to be hard-coded. This both restricts the total branching fraction coverage of the algorithm and relies on intuition to decide which decay processes to reconstruct. In this work we introduce a method for learning which processes to reconstruct and how to reconstruct them from example using graph neural networks.