Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube
IceCube, a cubic-kilometer array of optical sensors built to detect atmospheric and astrophysical neutrinos between 1 GeV and 1 PeV, is deployed 1.45 km to 2.45 km below the surface of the ice sheet at the South Pole. The classification and reconstruction of events from the in-ice detectors play a c...
Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
---|---|
Format: | Report |
Language: | English |
Published: |
2022
|
Subjects: | |
Online Access: | https://bib-pubdb1.desy.de/record/482414 https://bib-pubdb1.desy.de/search?p=id:%22PUBDB-2022-04763%22 |
id |
ftdesyvdb:oai:bib-pubdb1.desy.de:482414 |
---|---|
record_format |
openpolar |
spelling |
ftdesyvdb:oai:bib-pubdb1.desy.de:482414 2023-05-15T16:41:35+02:00 Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube Abbasi, R. Ackermann, Markus Andeen, K. Friedman, E. Fritz, A. Fürst, P. Gaisser, T. K. Gallagher, J. Ganster, E. Garcia, A. Garrappa, S. Gerhardt, L. Ghadimi, A. Anderson, T. Glaser, C. Glauch, T. Glüsenkamp, T. Goehlke, N. Gonzalez, J. G. Goswami, S. Grant, D. Gray, S. J. Grégoire, T. Griswold, S. Anton, G. Günther, C. Gutjahr, P. Haack, C. Hallgren, A. Halliday, R. Halve, L. Halzen, F. Hamdaoui, H. Minh, M. Ha Hanson, K. Argüelles, C. Hardin, J. Harnisch, A. A. Hatch, P. Haungs, A. Helbing, K. Hellrung, J. Henningsen, F. Heuermann, L. Hickford, S. Hill, C. Ashida, Y. Hill, G. C. Hoffman, K. D. Hoshina, K. DE 2022 https://bib-pubdb1.desy.de/record/482414 https://bib-pubdb1.desy.de/search?p=id:%22PUBDB-2022-04763%22 eng eng info:eu-repo/semantics/altIdentifier/doi/10.3204/PUBDB-2022-04763 info:eu-repo/semantics/altIdentifier/arxiv/arXiv:2209.03042 https://bib-pubdb1.desy.de/record/482414 https://bib-pubdb1.desy.de/search?p=id:%22PUBDB-2022-04763%22 info:eu-repo/semantics/openAccess doi:10.3204/PUBDB-2022-04763 detector geometry neutrino energy low optical IceCube GeV efficiency ice neural network atmosphere cosmic radiation absorption photon surface on-line pole trigger background cloud resolution scattering info:eu-repo/semantics/preprint info:eu-repo/semantics/publishedVersion 2022 ftdesyvdb https://doi.org/10.3204/PUBDB-2022-04763 2023-02-20T00:16:00Z IceCube, a cubic-kilometer array of optical sensors built to detect atmospheric and astrophysical neutrinos between 1 GeV and 1 PeV, is deployed 1.45 km to 2.45 km below the surface of the ice sheet at the South Pole. The classification and reconstruction of events from the in-ice detectors play a central role in the analysis of data from IceCube. Reconstructing and classifying events is a challenge due to the irregular detector geometry, inhomogeneous scattering and absorption of light in the ice and, below 100 GeV, the relatively low number of signal photons produced per event. To address this challenge, it is possible to represent IceCube events as point cloud graphs and use a Graph Neural Network (GNN) as the classification and reconstruction method. The GNN is capable of distinguishing neutrino events from cosmic-ray backgrounds, classifying different neutrino event types, and reconstructing the deposited energy, direction and interaction vertex. Based on simulation, we provide a comparison in the 1-100 GeV energy range to the current state-of-the-art maximum likelihood techniques used in current IceCube analyses, including the effects of known systematic uncertainties. For neutrino event classification, the GNN increases the signal efficiency by 18% at a fixed false positive rate (FPR), compared to current IceCube methods. Alternatively, the GNN offers a reduction of the FPR by over a factor 8 (to below half a percent) at a fixed signal efficiency. For the reconstruction of energy, direction, and interaction vertex, the resolution improves by an average of 13%-20% compared to current maximum likelihood techniques in the energy range of 1-30 GeV. The GNN, when run on a GPU, is capable of processing IceCube events at a rate nearly double of the median IceCube trigger rate of 2.7 kHz, which opens the possibility of using low energy neutrinos in online searches for transient events. Report Ice Sheet South pole DESY Publication Database (PUBDB) South Pole |
institution |
Open Polar |
collection |
DESY Publication Database (PUBDB) |
op_collection_id |
ftdesyvdb |
language |
English |
topic |
detector geometry neutrino energy low optical IceCube GeV efficiency ice neural network atmosphere cosmic radiation absorption photon surface on-line pole trigger background cloud resolution scattering |
spellingShingle |
detector geometry neutrino energy low optical IceCube GeV efficiency ice neural network atmosphere cosmic radiation absorption photon surface on-line pole trigger background cloud resolution scattering Abbasi, R. Ackermann, Markus Andeen, K. Friedman, E. Fritz, A. Fürst, P. Gaisser, T. K. Gallagher, J. Ganster, E. Garcia, A. Garrappa, S. Gerhardt, L. Ghadimi, A. Anderson, T. Glaser, C. Glauch, T. Glüsenkamp, T. Goehlke, N. Gonzalez, J. G. Goswami, S. Grant, D. Gray, S. J. Grégoire, T. Griswold, S. Anton, G. Günther, C. Gutjahr, P. Haack, C. Hallgren, A. Halliday, R. Halve, L. Halzen, F. Hamdaoui, H. Minh, M. Ha Hanson, K. Argüelles, C. Hardin, J. Harnisch, A. A. Hatch, P. Haungs, A. Helbing, K. Hellrung, J. Henningsen, F. Heuermann, L. Hickford, S. Hill, C. Ashida, Y. Hill, G. C. Hoffman, K. D. Hoshina, K. Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube |
topic_facet |
detector geometry neutrino energy low optical IceCube GeV efficiency ice neural network atmosphere cosmic radiation absorption photon surface on-line pole trigger background cloud resolution scattering |
description |
IceCube, a cubic-kilometer array of optical sensors built to detect atmospheric and astrophysical neutrinos between 1 GeV and 1 PeV, is deployed 1.45 km to 2.45 km below the surface of the ice sheet at the South Pole. The classification and reconstruction of events from the in-ice detectors play a central role in the analysis of data from IceCube. Reconstructing and classifying events is a challenge due to the irregular detector geometry, inhomogeneous scattering and absorption of light in the ice and, below 100 GeV, the relatively low number of signal photons produced per event. To address this challenge, it is possible to represent IceCube events as point cloud graphs and use a Graph Neural Network (GNN) as the classification and reconstruction method. The GNN is capable of distinguishing neutrino events from cosmic-ray backgrounds, classifying different neutrino event types, and reconstructing the deposited energy, direction and interaction vertex. Based on simulation, we provide a comparison in the 1-100 GeV energy range to the current state-of-the-art maximum likelihood techniques used in current IceCube analyses, including the effects of known systematic uncertainties. For neutrino event classification, the GNN increases the signal efficiency by 18% at a fixed false positive rate (FPR), compared to current IceCube methods. Alternatively, the GNN offers a reduction of the FPR by over a factor 8 (to below half a percent) at a fixed signal efficiency. For the reconstruction of energy, direction, and interaction vertex, the resolution improves by an average of 13%-20% compared to current maximum likelihood techniques in the energy range of 1-30 GeV. The GNN, when run on a GPU, is capable of processing IceCube events at a rate nearly double of the median IceCube trigger rate of 2.7 kHz, which opens the possibility of using low energy neutrinos in online searches for transient events. |
format |
Report |
author |
Abbasi, R. Ackermann, Markus Andeen, K. Friedman, E. Fritz, A. Fürst, P. Gaisser, T. K. Gallagher, J. Ganster, E. Garcia, A. Garrappa, S. Gerhardt, L. Ghadimi, A. Anderson, T. Glaser, C. Glauch, T. Glüsenkamp, T. Goehlke, N. Gonzalez, J. G. Goswami, S. Grant, D. Gray, S. J. Grégoire, T. Griswold, S. Anton, G. Günther, C. Gutjahr, P. Haack, C. Hallgren, A. Halliday, R. Halve, L. Halzen, F. Hamdaoui, H. Minh, M. Ha Hanson, K. Argüelles, C. Hardin, J. Harnisch, A. A. Hatch, P. Haungs, A. Helbing, K. Hellrung, J. Henningsen, F. Heuermann, L. Hickford, S. Hill, C. Ashida, Y. Hill, G. C. Hoffman, K. D. Hoshina, K. |
author_facet |
Abbasi, R. Ackermann, Markus Andeen, K. Friedman, E. Fritz, A. Fürst, P. Gaisser, T. K. Gallagher, J. Ganster, E. Garcia, A. Garrappa, S. Gerhardt, L. Ghadimi, A. Anderson, T. Glaser, C. Glauch, T. Glüsenkamp, T. Goehlke, N. Gonzalez, J. G. Goswami, S. Grant, D. Gray, S. J. Grégoire, T. Griswold, S. Anton, G. Günther, C. Gutjahr, P. Haack, C. Hallgren, A. Halliday, R. Halve, L. Halzen, F. Hamdaoui, H. Minh, M. Ha Hanson, K. Argüelles, C. Hardin, J. Harnisch, A. A. Hatch, P. Haungs, A. Helbing, K. Hellrung, J. Henningsen, F. Heuermann, L. Hickford, S. Hill, C. Ashida, Y. Hill, G. C. Hoffman, K. D. Hoshina, K. |
author_sort |
Abbasi, R. |
title |
Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube |
title_short |
Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube |
title_full |
Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube |
title_fullStr |
Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube |
title_full_unstemmed |
Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube |
title_sort |
graph neural networks for low-energy event classification & reconstruction in icecube |
publishDate |
2022 |
url |
https://bib-pubdb1.desy.de/record/482414 https://bib-pubdb1.desy.de/search?p=id:%22PUBDB-2022-04763%22 |
op_coverage |
DE |
geographic |
South Pole |
geographic_facet |
South Pole |
genre |
Ice Sheet South pole |
genre_facet |
Ice Sheet South pole |
op_source |
doi:10.3204/PUBDB-2022-04763 |
op_relation |
info:eu-repo/semantics/altIdentifier/doi/10.3204/PUBDB-2022-04763 info:eu-repo/semantics/altIdentifier/arxiv/arXiv:2209.03042 https://bib-pubdb1.desy.de/record/482414 https://bib-pubdb1.desy.de/search?p=id:%22PUBDB-2022-04763%22 |
op_rights |
info:eu-repo/semantics/openAccess |
op_doi |
https://doi.org/10.3204/PUBDB-2022-04763 |
_version_ |
1766032031097028608 |