Vertex and Energy Reconstruction in JUNO with Machine Learning Methods

Abstract

The Jiangmen Underground Neutrino Observatory (JUNO) is an experiment designed to study neutrino oscillations. Determination of neutrino mass ordering and precise measurement of neutrino oscillation parameters ${\sin }^{2}2{\theta }_{12}$, $\Delta m_{21}^2$ and $\Delta m_{32}^2$ are the main goals of the experiment. A rich physical program beyond the oscillation analysis is also foreseen. The ability to accurately reconstruct particle interaction events in JUNO is of great importance for the success of the experiment. In this work we present a few machine learning approaches applied to the vertex and the energy reconstruction. Multiple models and architectures were compared and studied, including Boosted Decision Trees (BDT), Deep Neural Networks (DNN), a few kinds of Convolution Neural Networks (CNN), based on ResNet and VGG, and a Graph Neural Network based on DeepSphere. Based on a study, carried out using the dataset, generated by the official JUNO software, we demonstrate that machine learning approaches achieve the necessary level of accuracy for reaching the physical goals of JUNO: ${\sigma }_E=3\%$ at $E_{\text{vis}}=1\text{MeV}$ for the energy and ${\sigma }_{x,y,z}=10\text{cm}$ at $E_{\text{vis}}=1\text{MeV}$ for the position.

arXiv comment30 pages, 21 figures

References in corpus (27)

• Adam: A Method for Stochastic Optimization
• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Scikit-learn: Machine Learning in Python
• XGBoost: A Scalable Tree Boosting System
• Convolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?
• Unsupervised Domain Adaptation by Backpropagation
• Deep Learning using Rectified Linear Units (ReLU)
• SGDR: Stochastic Gradient Descent with Warm Restarts
• Searching for Exotic Particles in High-Energy Physics with Deep Learning
• Neutrino Physics with JUNO
• A disciplined approach to neural network hyper-parameters: Part 1 -- learning rate, batch size, momentum, and weight decay
• Layer Normalization
• Adversarial Discriminative Domain Adaptation
• Deep Learning and its Application to LHC Physics
• JUNO Conceptual Design Report
• Jet Flavor Classification in High-Energy Physics with Deep Neural Networks
• DeepSphere: Efficient spherical Convolutional Neural Network with HEALPix sampling for cosmological applications
• Stochastic Gradient Descent as Approximate Bayesian Inference
• Calibration Strategy of the JUNO Experiment
• The HEALPix Primer
• Graph Neural Networks in TensorFlow and Keras with Spektral
• Event vertex and time reconstruction in large volume liquid scintillator detector
• Convolutional Neural Networks at Constrained Time Cost
• A ROOT Based Event Display Software for JUNO
• A method of detector and event visualization with Unity in JUNO
• Time and space reconstruction in optical, non-imaging, scintillator-based particle detectors
• A method for sharing dynamic geometry information in studies on liquid-based detectors

Cited by in corpus (15)

• Machine Learning in Nuclear Physics
• Method for detector description transformation to Unity and application in BESIII
• Applications and Techniques for Fast Machine Learning in Science
• Mass Testing and Characterization of 20-inch PMTs for JUNO
• Radioactivity control strategy for the JUNO detector
• Simulation Software of the JUNO Experiment
• Visualization for physics analysis improvement and applications in BESIII
• Machine learning method for ${}^{12}$C event classification and reconstruction in the active target time-projection chamber
• Energy reconstruction for large liquid scintillator detectors with machine learning techniques: aggregated features approach
• Distinctive nuclear signatures of low-energy atmospheric neutrinos
• Artificial intelligence for improved fitting of trajectories of elementary particles in inhomogeneous dense materials immersed in a magnetic field
• A ROOT based detector geometry and event visualization system for JUNO-TAO
• Interpretable machine learning approach for electron antineutrino selection in a large liquid scintillator detector
• First-principle event reconstruction by time-charge readouts for the Taishan Antineutrino Observatory
• Individual Neutrino Masses From a Supernova