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

AKPIK: Arbeitskreis Physik, moderne Informationstechnologie und Künstliche Intelligenz

AKPIK 2: Machine-learning methods and computing in particle physics

AKPIK 2.3: Vortrag

Dienstag, 26. März 2019, 16:20–16:30, H10

Analysis of GERDA detector surface events with deep learning algorithms — •Péter Kicsiny for the GERDA collaboration — Max-Planck-Institut für Physik, München, Deutschland

The Gerda experiment searches for the neutrinoless double beta (0νββ) decay of the ⁷⁶Ge in high purity germanium detectors enriched in this isotope. The detectors are operated in liquid argon. A primary background source around the Q-value of the 0νββ decay (Q_ββ=2039 keV) are β-decays of ⁴²K resulting from the contamination of natural argon with the isotope ⁴²Ar. The β-particles from the decay deposit their energy close to the detector surface. The rejection of these events is currently performed by a one parameter cut based on the current pulse amplitude divided by the total energy. In order to identify surface β-events separated from Compton scattered γ-events, more sophisticated methods are investigated. Artificial neural networks with advanced deep learning architectures are becoming more efficient in such classification tasks. Preliminary results on surface event classification using deep learning algorithms will be discussed using training data from ³⁹Ar β-decays present in the background spectrum at low energies (Q_β=565 keV).