Speaker
Thorsten Glüsenkamp
(Uppsala University)
Description
IceCube is a cubic-kilometre high-energy neutrino observatory at the South Pole. By now, it has accumulated more than 10 years of data and has started the field of high-energy neutrino astronomy with the detection of diffuse astrophysical neutrinos and evidence for a few point sources. While more data is still taken, the existing 10+ years of data is also constantly re-analyzed with better analysis techniques. However, because of data complexity and systematic uncertanties, classical statistical methods have reached their limit in extracting information. Recently, deep learning techniques have been shown outperform in various areas of reconstruction, and we highlight challenges and prospects of these new tools with some examples.
Primary author
Thorsten Glüsenkamp
(Uppsala University)
