Aachen 2019 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
T: Fachverband Teilchenphysik
T 4: Deep Learning I
T 4.4: Vortrag
Montag, 25. März 2019, 16:45–17:00, H06
Validation of a Deep Neural Network Based Flavor Tagging Algorithm at Belle and Belle II — •Jochen Gemmler, Florian Bernlochner, Michael Feindt, and Pablo Goldenzweig for the Belle 2 collaboration — ETP, KIT, Karlsruhe
Measurements of time dependent CP violation will be one of the key tasks of the Belle II experiment, which is located at the SuperKEKB collider in Tsukuba, Japan. Via electron-positron collisions, neutral B mesons are produced in an entangled state. For CP violation measurements, it is crucial to infer the flavor of the accompanying Btag meson from its final states, exploiting the flavor specific decay topologies. This process is referred to as flavor tagging and the presented approach utilizes a Deep Neural Network (DNN) for this classification.
This talk shows current results of the validation of the DNN based approach on the full Belle dataset, which contains 772 Million B-Meson pairs, using the software framework of Belle II.