Ponente
Descripción
Hyper-Kamiokande (HK) is a next-generation neutrino experiment in Japan featuring a large water-Cherenkov detector. It aims to tackle key questions in fundamental physics, including precise neutrino oscillation measurements, the study of astrophysical neutrinos (like those from supernovae and the Diffuse Supernova Neutrino Background), and searches for proton and exotic nucleon decays.
Building on over a decade of experience from its predecessor, Super-Kamiokande, HK leverages neutron tagging to enhance sensitivity in these investigations. Neutrons from interactions in HK thermalize and are captured by hydrogen, emitting a faint 2.2 MeV photon—too weak for the standard trigger, requiring post-event PMT scan analysis.
This study introduces a neural network-based method to identify PMTs likely to detect this delayed signal. The approach significantly improves neutron signal selection, boosting efficiency from 58% to 75% over traditional threshold-based methods, while also enhancing candidate purity and aiding in subsequent signal reconstruction.
