Fast low energy reconstruction using Convolutional Neural Networks

Abstract IceCube is a Cherenkov detector instrumenting over a cubic kilometer of glacial ice deep under the surface of the South Pole. The DeepCore sub-detector lowers the detection energy threshold to a few GeV, enabling the precise measurements of neutrino oscillation parameters with atmospheric neutrinos. The reconstruction of neutrino interactions inside the detector is essential in studying neutrino oscillations. It is particularly challenging to reconstruct sub-100 GeV events with the IceCube detectors due to the relatively sparse detection units and detection medium. Convolutional neural networks (CNNs) are broadly used in physics experiments for both classification and regression purposes. This paper discusses the CNNs developed and employed for the latest IceCube-DeepCore oscillation measurements [1]. These CNNs estimate various properties of the detected neutrinos, such as their energy, direction of arrival, interaction vertex position, flavor-related signature, and are also used for background classification.

Authors

Institutions

Publication Details

Journal
Journal of Instrumentation
Published
2026-02-01
DOI
https://doi.org/10.1088/1748-0221/21/02/p02020
Primary Topic
Astrophysics and Cosmic Phenomena
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Fast low energy reconstruction using Convolutional Neural Networks

S.R. Klein, B.A. Clark, J.M. Conrad, J.P. Yanez et al.
Journal of Instrumentation
Astrophysics and Cosmic Phenomena
article

Fast low energy reconstruction using Convolutional Neural Networks

S.R. Klein, B.A. Clark, J.M. Conrad, J.P. Yanez, DJ Koskinen, F.G. Schröder, SK Agarwalla, C. Chen, R.T. Burley, Pa Evenson, A.R. Fazely, J.J. DeLaunay, Y.T. Liu, K.D de Vries, S.W. Barwick, Gm Spiczak, A. Y. Wen, Gc Hill, C.D. Rho, MU Nisa, MJ Larson, GW Sullivan, R.W Moore, P.A. Sevle Myhr, K.D Hoffman, E.J Roberts, S.J Gray, I.C Maris, K.M Groth, A.A Harnisch, J.A Aguilar, G.H Collin, A.K Upadhyay, S.N Axani, J.P Lazar, N.M Amin, E.N Paudel, K.L Fan, J.Y. Book Motzkin, D.Z Besson, J.M Alameddine, A. Kochocki, L. Gerhardt, L. Draper, D. Seckel, D. Kang, J. Carpio, S. Athanasiadou, D. Soldin, K. Leonard DeHolton, S. Fukami, T. Ehrhardt, J. Mitchell, F. Lauber, J. Gallagher, P. Behrens, T. Kozynets, P. Gutjahr, S. Tilav, K. Fang, N. Feigl, M. Seikh, R. Naab, J. Becker Tjus, B. Brinson, S. DiKerby, H. Pandya, K. Rawlins, N. Kurahashi, W. Rhode, A. Hallgren, R. Babu, V. Poojyam, J. Osborn, A. Coleman, K. Noda, R. Hussain, N. Krieger, P. Sampathkumar, L. Hennig, J. Braun, A. Novikov, H. Kimku, B. Schlüter, K. Hultqvist, A. Haungs, E. Genton, Y. Ashida, A. Connolly, M. Neumann, A. Mand, B. Riedel, T. Kontrimas, A. Rehman, H. Niederhausen, S. Mancina, J. Thwaites, M. Venugopal, V. Basu, M. Ravn, S. Zhang, K. Upshaw, A. Fedynitch, C. Raab, C. Lagunas Gualda, A. Granados, H. Schieler, L. Schlickmann, J.G. Gonzalez, F. Halzen, J. Stachurska, T. DeYoung, N. Schmeisser, S. Rodan, R. Abbasi, C. Tönnis, R. Hewett, S. Ter-Antonyan, M. Santander, P. Weigel, L. Molchany, A. Hidvegi, J. Rack-Helleis, L. Neste, N. Shimizu, K. Andeen, K. Hoshina, J. Adams, K. Farrag, M. Seo, J. Beise, S. Jain, R. Orsoe, S. Toscano, A. Olivas, J. Häußler, J. Necker, S. Hori, N. Whitehorn, A. Karle, J. Liao, J. Saffer, L. Heuermann, I. Reistroffer, A. Eimer, A. Søgaard, J. Soedingrekso, J. Micallef, V. Parrish, C. Wendt, L. Bloom, J. Kiryluk, D. Durnford, L. Pyras, N. van Eijndhoven, A. Obertacke Pollmann, C. Hill, A. Vaidyanathan, Y. Morii, B. Skrzypek, G. Sommani, M. Kowalski, C. Boscolo Meneguolo, Y. Popovych, K. Carloni, A. Noell, C. Kopper, E. Manao, M. Wolf, A. Franckowiak, A. Rifaie, D. Grant, E. Kun, Z. Marka, N. Lad, P. Eller, X. Bai, T. Ruhe, R. Maruyama, P. Coleman, M. Ackermann, M. Vereecken, M. Handt, A. Rosted, A. Khatee Zathul, N. Chau, L. Schumacher, A. Pizzuto, J. Willison, K. Meagher, X.W. Xu, C. Argüelles, A. Mosbrugger, A. Parenti, E. O'Sullivan, A. Manish Kumar, F. Varsi, D.F. Cowen, D. Berley, S. Choi, S. Chattopadhyay, A. Ghadimi, V. O'Dell, K. Kruiswijk, P. Soldin, M. Liubarska, T. Karg, J. Hardin, G. Wrede, M. Ha Minh, N. Kamp, M. Schaufel, C. Finley, A. Tavira García, W. Kang, M. D. Durán García, D. Delgado, L. Witthaus, C.H. Wiebusch, M. Ahlers, M. Jacquart, F. Lucarelli, C. Glaser, L. Paul, D. Guevel, M. A. S. da SILVA, X. Kang, C. Love, F. Schlüter, M. Zimmerman, R. Corley, S. Sclafani, Z. Brisson-Tsavoussis, C. Haack, J. Veitch-Michaelis, A. Kheirandish, C. De Clercq, O. Botner, J. Hellrung, H. Dujmovic, A. Zegarelli, W.G. Thompson, E. Warrick, T. Van Eeden, S. BenZvi, M.A. DuVernois, R. Hmaid, S. Yoshida, S. Blot, W. Esmail, D. Fox, A. Sandrock, E. Ganster, A. Pontén, S. Robertson, P. Fürst, A. Thakuri, A. Weindl, J.L. Kelley, Z. Rechav, B. Pries, C. Pérez de los Heros, S. Yu, C. Klein, S. Goswami, A. Desai, M. Prado Rodriguez, O. Janik, Y. Lyu, L. Marten, D. Salazar-Gallegos, N. Valtonen-Mattila, P. Zilberman, C. Ha, G. Garg, M. Dittmer, D. Veske, T. Stuttard, D. Chirkin, L. Fischer, L. Seen, F. Mayhew, F. Bontempo, L. Hamacher, J. Villarreal, J. van Santen, L. Eidenschink, A. Kappes, L. Lu, Y. Makino, R. Snihur, T. Glüsenkamp, Y. Merckx, S. Seunarine, J. Werthebach, M. Thiesmeyer, M. Karl, L. Ruohan, F. Yu, J. Krishnamoorthi, L. Kardum, D.R. Williams, T. Schmidt, U. Naumann, S. Reusch, M. Khanal, J. Böttcher, P. Savina, G. de Wasseige, E. Resconi, J.J. Beatty, S. Griffin, R. Shah, S. Schwirn, P. Desiati, G.T. Przybylski, C. Walck, J. Vara, S. Verpoest, F. McNally, S. Marka, M. Nakos, A. Domi, A. Wang, D. Butterfield, Y. Kobayashi, C. Rott, K. Hanson, K. Hymon, R. Engel, T. Stürwald, E. Yildizci, M. Baricevic, T. Stezelberger, J. Madsen, A. Medina, K.B.M. Mahn, S. Mechbal, N. Rad, C. Spiering, T. Pernice, R. Young, M. Jansson, E. Bernardini, T. Stanev, P. Hatch, M. Kauer, I. Taboada, H. Kolanoski, S. Hickford, A. Ishihara, T. Yuan, M. Stamatikos, T. Montaruli, C. Bellenghi, J. Vandenbroucke, Z. T. Chen, R. Procter-Murphy, A. Vijai, J. Savelberg, M. Meier, W. Hou, R. Bay, K. Helbing, A. Terliuk, P. Zhelnin, M. Plum, M. Moulai, N. Heyer, P. Koundal, W. Iwakiri, W. Luszczak, A. Trettin, D. Tosi, R. Koirala, L. Dueser, M. Hostert, M. Rongen, P. Schaile, T. Huber, D. Elsässer, M. Weyrauch, E. Krupczak, T. Krishnan, L. Halve, L. Merten, S. Böser, E. Ellinger, K. Dutta, S. Sarkar, A. Katil, L. Köpke, W. Marie Sainte, J. Peterson, E. Magnus, J. Evans, C. Weaver, S. Shefali, A. Leszczyńska, Z H Zhang, S. Bash, J.C. Díaz-Vélez, S. Schindler, V. Palusova, M. Jeong, H. Erpenbeck, N. Park, S. Vergara Carrasco, K. Tollefson, I. Martinez-Soler, M.A. Unland Elorrieta, A.Balagopal V., M. Jin, M. Lamoureux, R. Morse, I. Safa, C. Spannfellner, S. Griswold, C. Günther, S. Deng, J.V. Mead, E. Blaufuss, T. K. Mukherjee
article en

Abstract

Abstract IceCube is a Cherenkov detector instrumenting over a cubic kilometer of glacial ice deep under the surface of the South Pole. The DeepCore sub-detector lowers the detection energy threshold to a few GeV, enabling the precise measurements of neutrino oscillation parameters with atmospheric neutrinos. The reconstruction of neutrino interactions inside the detector is essential in studying neutrino oscillations. It is particularly challenging to reconstruct sub-100 GeV events with the IceCube detectors due to the relatively sparse detection units and detection medium. Convolutional neural networks (CNNs) are broadly used in physics experiments for both classification and regression purposes. This paper discusses the CNNs developed and employed for the latest IceCube-DeepCore oscillation measurements [1]. These CNNs estimate various properties of the detected neutrinos, such as their energy, direction of arrival, interaction vertex position, flavor-related signature, and are also used for background classification.

Journal of InstrumentationVol. 21(02)
Karlsruhe Institute of Technology (DE), Marquette University (US), University of Geneva (CH), Uppsala University (SE), University of Copenhagen (DK), University of Wisconsin System (US), Georgia Institute of Technology (US), Pennsylvania State University (US), Université Libre de Bruxelles (BE), University of Nevada, Las Vegas (US), Vrije Universiteit Brussel (BE), University of Wisconsin–Madison (US), University of Padua (IT), University of Wisconsin–River Falls (US), Southern University and Agricultural and Mechanical College (US), University of Kansas (US), University of Alaska Anchorage (US), Lawrence Berkeley National Laboratory (US), Mercer University (US), University of Alberta (CA), Chiba University (JP), Stockholm University (SE), University of Wuppertal (DE), University of Alabama (US), Simon Fraser University (CA), Friedrich-Alexander-Universität Erlangen-Nürnberg (DE), South Dakota School of Mines and Technology (US), University of Canterbury (NZ), Loyola University Chicago (US), Johannes Gutenberg University Mainz (DE), TU Dortmund University (DE), University of California, Irvine (US), Queen's University (CA), University of Utah (US), University of Münster (DE), Deutsches Elektronen-Synchrotron DESY (DE), Ghent University (BE), Yale University (US), Humboldt-Universität zu Berlin (DE), University of Oxford (GB), Institute of Particle Physics (CA), Institute of Physics, Academia Sinica (TW), The Ohio State University (US), University of Rochester (US), The University of Adelaide (AU), Stony Brook University (US), Technical University of Munich (DE), Massachusetts Institute of Technology (US), University of Maryland, College Park (US), Chung-Ang University (KR), Drexel University (US), Southern University (BD), Columbia University (US), Sungkyunkwan University (KR), University of Delaware (US), Michigan State University (US), RWTH Aachen University (DE), Ruhr University Bochum (DE), University of California, Berkeley (US), UCLouvain (BE)
Affordable and clean energy
Openalex Percentile: Top 57%
Astrophysics and Cosmic Phenomena
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.