Abstract
The KM3NeT Collaboration is building a network of underwater Cherenkov telescopes at two sites in the Mediterranean Sea, with the main goals of investigating astrophysical sources of high-energy neutrinos (ARCA) and of determining the neutrino mass hierarchy (ORCA). Various Machine Learning techniques, such as Random Forests, BDTs, Shallow and Deep Networks are being used for diverse tasks, such as event-type and particle identification, energy/direction estimation, source identification, signal/background discrimination and data analysis, with sound results as well as promising research paths. The main focus of this work is the application of Convolutional Neural Network models to the tasks of neutrino interaction classification, as well as the estimation of energy and direction of the propagating particles. The performances are also compared to those of the standard reconstruction algorithms used in the Collaboration.
Highlights
KM3NeT is a multi-km3 network of neutrino telescopes [1] designed with the main goals of identifying sources of cosmic neutrinos [2] and establishing the neutrino mass hierarchy [3]
Convolutional Neural Networks (CNN) [4] are applied to the problem of identifying and studying neutrino interactions in order to classify KM3NeT events according to their topological features
Results obtained from the Deep Learning models are compared to the KM3NeT official track reconstruction algorithm for νμCC events, and/or other Machine Learning-based solutions
Summary
KM3NeT is a multi-km network of neutrino telescopes [1] designed with the main goals of identifying sources of cosmic neutrinos [2] and establishing the neutrino mass hierarchy [3]. The detector is composed of a three-dimensional array of photomultipliers (PMT), hosted in spherical structures called Digital Optical Modules (DOM). In this setting, neutrinos are indirectly detected by collecting the Cherenkov light emitted by the propagation in the sea water of the particles produced in the interaction. Convolutional Neural Networks (CNN) [4] are applied to the problem of identifying and studying neutrino interactions in order to classify KM3NeT events according to their topological features (i.e. space-time distribution of photomultiplier signals). Results obtained from the Deep Learning models are compared to the KM3NeT official track reconstruction algorithm for νμCC events, and/or other Machine Learning-based solutions
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
