Gavrikov, A. (2024) Energy reconstruction with machine learning techniques in JUNO. Il nuovo cimento C, 47 (6). pp. 1-4. ISSN 1826-9885
|
Text
ncc13148.pdf - Published Version Download (194kB) | Preview |
Abstract
The Jiangmen Underground Neutrino Observatory (JUNO) is a multipurpose liquid scintillator neutrino experiment under construction located in China. Although the main source of neutrinos in JUNO is two nuclear power plants located about 52.5 km away from the experiment, it will also be able to study solar and atmospheric neutrinos, geoneutrinos and neutrinos coming from supernovae. The determination of the neutrino mass ordering (NMO) with 3-4σ in 6 years is the main goal of the experiment. Moreover, another JUNO’s important aim is to measure neutrino oscillation parameters sin2 θ12,Δm2 21,Δm2 31 with sub-percent precision. The central detector of JUNO is an acrylic sphere filled with 20 kt of liquid-scintillator (LS) surrounded by 17612 20-inch photomultiplier tubes (PMTs) and 25600 3-inch PMTs, providing ∼78% coverage of the detector sphere. Thanks to the almost complete coverage of the sphere by the PMTs array, as well as high light yield leads to an unprecedented, for LS-based experiments, energy resolution of 3% at 1 MeV. Due to the need to take into account various effects, including the non-linearity of the energy response and the detector’s spatial non-uniformity, event energy reconstruction is not a straightforward task. In this study, energy reconstruction for reactor neutrino events with machine learning (ML) techniques is presented. The following two models are used: Boosted Decision Trees and Fully Connected Deep Neural Network. The models are trained on aggregated features extracted from charge and time information on PMTs.
| Item Type: | Article |
|---|---|
| Subjects: | 500 Scienze naturali e Matematica > 530 Fisica |
| Depositing User: | Marina Spanti |
| Date Deposited: | 10 Feb 2025 13:53 |
| Last Modified: | 10 Feb 2025 13:53 |
| URI: | http://eprints.bice.rm.cnr.it/id/eprint/23400 |
Actions (login required)
 |
View Item |
