Bonn 2020 – wissenschaftliches Programm
Die DPG-Frühjahrstagung in Bonn musste abgesagt werden! Lesen Sie mehr ...
T 81.7: Vortrag
Donnerstag, 2. April 2020, 18:00–18:15, H-1.004
Event Selection of Muons with Recurrent Neural Networks in IceCube — •Gerrit Wrede, Thorsten Glüsenkamp, and Gisela Anton for the IceCube collaboration — Erlangen Centre for Astroparticle Physics (ECAP), University Erlangen-Nürnberg, Germany
Relativistic muons created by muon neutrinos offer a good angular resolution and are thus an ideal channel for the detection of point sources. However, IceCube also measures a major background of muons from cosmic rays. This background can be avoided by selecting upgoing events from the northern hemisphere, where the earth shields cosmic rays. It is essential to reject misreconstructed cosmic-ray events in the upgoing region while preserving as much signal as possible. We trained a recurrent neural network to reconstruct muon tracks with high accuracy. In this talk, the neural-network-based reconstruction will be presented and its ability to reject cosmic-ray muons will be compared with standard reconstructions.