Indico has been upgraded. It is back in production.

25 March 2022
H.C. Ørsted Institute
Europe/Copenhagen timezone
Abstract submission and Registration is now open!

Machine Learning In IceCube: A Graphical Approach

25 Mar 2022, 16:05
1h 40m
Hall (H.C.Ørsted Institute)

Hall

H.C.Ørsted Institute

Speakers

Mr Kaare Endrup Iversen (NBI) Leon B (NBI)

Description

The IceCube Detector is a neutrino telescope designed to study the cosmos from deep within the ice sheet of the south pole. A cubic kilometre of ice is used to detect the Cherenkov radiation from charged decay products of neutrinos. With thousands of triggered events every second, sophisticated methods are needed to process the data. Our use of Graphical Neural Networks provides an alternative reconstruction method to the reliable, but slow, current regime.
We present new analyses of stopped muons contained inside the detector, estimating the inelasticity of neutrino/anti-neutrino decays as well as using GNNs for noise cleaning in the forthcoming IceCube upgrade detector.

Field of study Computational Physics
Supervisor Troels C. Petersen

Primary authors

Presentation materials

There are no materials yet.