Deep Neural Networks for combined neutrino energy estimate with KM3NeT/ORCA6

• Published in: Proceedings of 38th International Cosmic Ray Conference — PoS(ICRC2023)
• Main Author: Peña Martínez, Santiago
• Other Authors: AstroParticule et Cosmologie (APC (UMR_7164)), Commissariat à l'énergie atomique et aux énergies alternatives (CEA)-Institut National de Physique Nucléaire et de Physique des Particules du CNRS (IN2P3)-Observatoire de Paris, Université Paris Sciences et Lettres (PSL)-Université Paris Sciences et Lettres (PSL)-Centre National de la Recherche Scientifique (CNRS)-Université Paris Cité (UPCité), KM3NeT
• Format: Conference Object
• Language: English
• Published: HAL CCSD 2023
• Subjects: atmosphere; KM3NeT; oscillation; sensitivity; neural network; neutrino: energy; [PHYS.ASTR]Physics [physics]/Astrophysics [astro-ph]; [PHYS.HEXP]Physics [physics]/High Energy Physics - Experiment [hep-ex]; Orca
• Online Access: https://hal.science/hal-04185320; https://doi.org/10.22323/1.444.1035

International audience KM3NeT/ORCA is a large-volume water-Cherenkov neutrino detector, currently under construction at the bottom of the Mediterranean Sea at a depth of 2450 meters. The main research goal ofORCA is the measurement of the neutrino mass ordering and the atmospheric neutrino oscillationparameters. Additionally, the detector is also sensitive to a wide variety of phenomena includingnon-standard neutrino interactions, sterile neutrinos, and neutrino decay.This contribution describes the use of a machine learning framework for building Deep NeuralNetworks (DNN) which combine multiple energy estimates to generate a more precise reconstructed neutrino energy. The model is optimized to improve the oscillation analysis based ona data sample of 433 kton-years of KM3NeT/ORCA with 6 detection units. The performanceof the model is evaluated by determining the sensitivity to oscillation parameters in comparisonwith the standard energy reconstruction method of maximizing a likelihood function. The resultsshow that the DNN is able to provide a better energy estimate with lower bias in the context ofoscillation analyses.