BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:UW-Madison-Physics-Events
BEGIN:VEVENT
SEQUENCE:5
UID:UW-Physics-Event-8199
DTSTART:20230217T180000Z
DTEND:20230217T200000Z
DTSTAMP:20260414T054106Z
LAST-MODIFIED:20230216T200813Z
LOCATION:2335 Sterling or https://uwmadison.zoom.us/j/5745114206?pwd=V
 0pzekNXZHhnZGF2UkU3UjRpcUFXdz09
SUMMARY:2D Convolutional Network for IceCube DeepCore Oscillation Stud
 ies\, Preliminary Exam\, Josh Peterson\, Physics Graduate Student
DESCRIPTION:IceCube DeepCore is an extension of the IceCube Neutrino O
 bservatory designed to measure GeV scale atmospheric neutrino interact
 ions. Neutrino reconstruction and classification tasks are especially 
 difficult at GeV scale energies in IceCube DeepCore due to sparse inst
 rumentation. Convolutional neural networks (CNNs) have been found to h
 ave better success at neutrino event reconstruction than likelihood-ba
 sed methods. I present a new CNN model that exploits time and depth tr
 anslational symmetry in IceCube DeepCore data and present the model’
 s performance\, specifically for neutrino flavor identification. This 
 new CNN model will be used for inelasticity reconstruction and/or matt
 er effect dependent neutrino oscillation studies.
URL:https://www.physics.wisc.edu/events/?id=8199
END:VEVENT
END:VCALENDAR