Identification of Patterns in Cosmic-Ray Arrival Directions using Dynamic Graph Convolutional Neural Networks
arXiv:2003.13038 · doi:10.1016/j.astropartphys.2020.102527
Abstract
We present a new approach for the identification of ultra-high energy cosmic rays from sources using dynamic graph convolutional neural networks. These networks are designed to handle sparsely arranged objects and to exploit their short- and long-range correlations. Our method searches for patterns in the arrival directions of cosmic rays, which are expected to result from coherent deflections in cosmic magnetic fields. The network discriminates astrophysical scenarios with source signatures from those with only isotropically distributed cosmic rays and allows for the identification of cosmic rays that belong to a deflection pattern. We use simulated astrophysical scenarios where the source density is the only free parameter to show how density limits can be derived. We apply this method to a public data set from the AGASA Observatory.
15 pages, 9 figures
References in corpus (56)
- Adam: A Method for Stochastic Optimization
- TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
- Semi-Supervised Classification with Graph Convolutional Networks
- HEALPix -- a Framework for High Resolution Discretization, and Fast Analysis of Data Distributed on the Sphere
- Cosmological Magnetic Fields: Their Generation, Evolution and Observation
- Observation of the suppression of the flux of cosmic rays above 4x10^19eV
- A New Model of the Galactic Magnetic Field
- Measurement of the energy spectrum of cosmic rays above 10^18 eV using the Pierre Auger Observatory
- ParticleNet: Jet Tagging via Particle Clouds
- Small-scale anisotropy of cosmic rays above 10^19eV observed with the Akeno Giant Air Shower Array
- Observation of a Large-scale Anisotropy in the Arrival Directions of Cosmic Rays above eV
- Depth of Maximum of Air-Shower Profiles at the Pierre Auger Observatory: Measurements at Energies above 10^17.8 eV
- Combined fit of spectrum and composition data as measured by the Pierre Auger Observatory
- The Galactic Magnetic Field
- CRPropa 3 - a Public Astrophysical Simulation Framework for Propagating Extraterrestrial Ultra-High Energy Particles
- Depth of Maximum of Air-Shower Profiles at the Pierre Auger Observatory: Composition Implications
- Representation Learning on Graphs: Methods and Applications
- Deriving global structure of the Galactic Magnetic Field from Faraday Rotation Measures of extragalactic sources
- Dynamic Graph CNN for Learning on Point Clouds
- The Cosmic Ray Energy Spectrum Observed with the Surface Detector of the Telescope Array Experiment
- Indication of anisotropy in arrival directions of ultra-high-energy cosmic rays through comparison to the flux pattern of extragalactic gamma-ray sources
- Search for Neutrinoless Double-Beta Decay with the Complete EXO-200 Dataset
- The Machine Learning Landscape of Top Taggers
- Ultra high energy cosmic rays and the large scale structure of the galactic magnetic field
- Evidence for Gamma-Ray Halos Around Active Galactic Nuclei and the First Measurement of Intergalactic Magnetic Fields
- DeepSphere: Efficient spherical Convolutional Neural Network with HEALPix sampling for cosmological applications
- Cosmic Ray Anisotropy Analysis with a Full-Sky Observatory
- Large-scale cosmic-ray anisotropies above 4 EeV measured by the Pierre Auger Observatory
- Pulsar Rotation Measures and Large-scale Magnetic Field Reversals in the Galactic Disk
- Mapping deflections of extragalactic Ultra-High Energy Cosmic Rays in magnetohydrodynamic simulations of the Local Universe
- Application of Deep Learning methods to analysis of Imaging Atmospheric Cherenkov Telescopes data
- Caltech Faint Galaxy Redshift Survey XIV: Galaxy Morphology in the HDF (North) and its Flanking Fields to z=1.2
- Lensing of ultra-high energy cosmic rays in turbulent magnetic fields
- CMB Constraints on a Stochastic Background of Primordial Magnetic Fields
- A Deep Learning-based Reconstruction of Cosmic Ray-induced Air Showers
- New constraints on modelling the random magnetic field of the MW
- Bounds on the density of sources of ultra-high energy cosmic rays from the Pierre Auger Observatory
- Propagation of Ultra High Energy Cosmic Rays in Extragalactic Magnetic Fields: A view from cosmological simulations
- Signatures of galactic magnetic lensing upon ultra high energy cosmic rays
- Ultrahigh Energy Nuclei in the Turbulent Galactic Magnetic Field
- Mapping UHECRs deflections through the turbulent galactic magnetic field with the latest RM data
- An upper limit on the strength of the extragalactic magnetic field from ultra-high-energy cosmic-ray anisotropy
- Search for signatures of magnetically-induced alignment in the arrival directions measured by the Pierre Auger Observatory
- Deep Neural Networks for Energy and Position Reconstruction in EXO-200
- Measurement of the cosmic ray spectrum above eV using inclined events detected with the Pierre Auger Observatory
- Ultrahigh Energy Nuclei in the Galactic Magnetic Field
- Source position reconstruction and constraints on the galactic magnetic field from ultra-high energy cosmic rays
- Updated AGASA event list above 4x10^19eV
- Do Ultrahigh Energy Cosmic Rays Come from Active Galactic Nuclei and Fermi -ray Sources?
- Deep learning techniques applied to the physics of extensive air showers
- Classification and Recovery of Radio Signals from Cosmic Ray Induced Air Showers with Deep Learning
- Search for patterns by combining cosmic-ray energy and arrival directions at the Pierre Auger Observatory
- A global autocorrelation study after the first Auger data: impact on the number density of UHECR sources
- Origins of Extragalactic Cosmic Ray Nuclei by Contracting Alignment Patterns induced in the Galactic Magnetic Field
- Searching for signals of magnetic lensing in ultra-high energy cosmic rays
- Prospects of detecting a large-scale anisotropy of ultra-high-energy cosmic rays from a nearby source with the K-EUSO orbital telescope
Cited by in corpus (7)
- Ultra-High-Energy Cosmic Rays: The Intersection of the Cosmic and Energy Frontiers
- Shared Data and Algorithms for Deep Learning in Fundamental Physics
- UHECR mass composition at highest energies from anisotropy of their arrival directions
- Application of Graph Networks to background rejection in Imaging Air Cherenkov Telescopes
- Deep learning techniques for Imaging Air Cherenkov Telescopes
- Application of Graph Networks to a wide-field Water-Cherenkov-based Gamma-Ray Observatory
- Towards extracting cosmic magnetic field structures from cosmic-ray arrival directions