Gießen 2024 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
HK: Fachverband Physik der Hadronen und Kerne
HK 72: Poster
HK 72.49: Poster
Donnerstag, 14. März 2024, 17:15–18:45, HBR 14: Foyer
Development of Machine Learning Algorithms to Optimise the Detection of Low-mass Dileptons — •Saket Sahu1, Johan Messchendorp2, and James Ritman1,2,3 for the HADES collaboration — 1Ruhr-Universität Bochum, Bochum, Germany — 2GSI Helmholtzzentrumfür Schwerionenforschung GmbH, Darmstadt, Germany — 3Forschungszentrum Jülich, Jülich, Germany
Radiative transitions and decays of hadrons provide valuable information on their electromagnetic structure. Particular, the usage of virtual photon (dileptons) is promising since it allows to extract observables, such as spin-density matrix elements (SDMEs), that are not accessible using real photons. The experimental challenges lie in the identification of (mostly) low-mass dilepton pairs and separating the physics channels of interest from bremsstrahlung and external conversion processes. The High Acceptance Di- Electron Spectrometer (HADES) at GSI Darmstadt is designed for an excellent e+/e− reconstruction in hadronic reactions. The current reconstruction algorithm fails to efficiently identify dilepton pairs with very small opening angles. Convolutional Neural Networks (CNN) are known to show great performance in image analysis and thus can be used for ring reconstruction. This poster outlines the analysis strategy for the SDME extraction based on recently taken data in proton-proton collisions with HADES, with an outlook on the implementation of the CNN for the ring reconstruction.
Keywords: Radiative Transition; Baryon Resonances; HADES; Machine Learning; Convolutional Neural Network