Speaker
Shuyang Deng
(RWTH Aachen University)
Description
The northern track event selection utilizes the event topology in the IceCube to select track-like up-going events, which are produced by high-energy muons induced by muon-neutrinos which penetrated the Earth and reach the detector from the Northern sky. The Efficiency of the selection is limited by the quality of event reconstructions, which might misidentify down-going muon tracks as up-going. To improve the efficiency of this selection method, we investigate the application of deep neuron network, which enables a selection without the prerequisite of event-reconstruction. The initial tests demonstrated an increasing efficiency in the event selection.
Primary author
Shuyang Deng
(RWTH Aachen University)
