Particle Discrimination via Deep Learning with JUNO

The JUNO detector is going to be a 20 kt liquid scintillator neutrino observatory, currently under construction near Kaiping, China, with a baseline of about 50km to two nuclear reactor plants. With its excellent energy resolution and large fiducial volume, it will be able to determine the neutrino mass hierarchy from their energy spectrum. The neutrinos are detected by measuring the signature of the inverse beta decay (IBD), which consists of a prompt positron- and a delayed neutron capture signal. Although this coincidence is well recognizable, there are still some background events left. Most of these backgrounds have of an electron component instead of a positron. Electron and positron signals are very similar in a liquid scintillator. The only difference is the missing annihilation for electrons. New developments in deep learning techniques give the possibility to distinguish the different event shapes. This talk focuses on a method to discriminate positrons and electrons via a neural network.