Extraction of Stopping Muons in IceCube Using Machine Learning

IceCube is a neutrino observatory located at the South Pole, consisting of digital optical modules that detect Cherenkov light emitted from charged particles traversing the Antarctic ice sheet. The most dominant source of background in the search for neutrinos are atmospheric muons produced in interactions of cosmic rays in the upper atmosphere. At a trigger rate of about 3000 Hz, they also provide a valuable source of information about cosmic rays. A special subset of atmospheric muon events are muon events that contain only muon tracks that end within the detector volume. These stopping muons exhibit features that have interesting implications for both cosmic-ray physics and calibration purposes. In order to extract a sample of stopping muons of high purity, a supervised machine learning approach is used. In this talk, an overview of the employed methods is given. The properties of the extracted sample are discussed, and an unfolding of the muon range spectrum is presented.