A convolutional neural network based cascade reconstruction for the IceCube Neutrino Observatory

@article{Abbasi2021ACN,
  title={A convolutional neural network based cascade reconstruction for the IceCube Neutrino Observatory},
  author={R. U. Abbasi and Markus Ackermann and J. H. Adams and Juan Antonio Aguilar and Markus Ahlers and Maximilian Ahrens and Cyril Martin Alispach and A. A. Alves and Najia Moureen Binte Amin and Rui An and Karen Andeen and Tyler Brooks Anderson and I. Ansseau and Gisela Anton and Carlos A. Arguelles and Spencer N. Axani and Xinhua Bai and V. A.Balagopal and A. Barbano and Steven W. Barwick and B. Bastian and V. Basu and V. Baum and S. Baur and Ryan Bay and James J. Beatty and Kurt H. Becker and Julia K. Becker Tjus and C. Bellenghi and Segev BenZvi and David Berley and Elisa Bernardini and David Z. Besson and Gary Binder and Daniel Bindig and Erik Blaufuss and S. Blot and Sebastian Boser and Olga Botner and Jakob Bottcher and {\'E}tienne Bourbeau and James Bourbeau and F. Bradascio and J. Braun and Stephanie Bron and Jannes Brostean-Kaiser and A. Burgman and Raffaela Busse and M. A. Campana and C. Chen and Dmitry Chirkin and S. Choi and B. A. Clark and Ken Clark and Lew Classen and A. Coleman and Gabriel H. Collin and Janet M. Conrad and Paul Coppin and P. Correa and D. F. Cowen and R. Cross and P. Dave and Catherine De Clercq and James DeLaunay and Hans Peter Dembinski and Kunal Deoskar and Simon De Ridder and Abhishek Desai and Paolo Desiati and Krijn D. de Vries and G. de Wasseige and Meike de With and Tyce DeYoung and S. Dharani and A. Diaz and J. C. D'iaz-V'elez and H. Dujmovic and Matthew Dunkman and Michael DuVernois and Emily Dvorak and Thomas Ehrhardt and Philipp Eller and Ralph Engel and J. J. Evans and Paul A. Evenson and Sam Fahey and Ali R. Fazely and Sebastian Fiedlschuster and A. T. Fienberg and Kirill Filimonov and Chad Finley and Leander Fischer and Derek B. Fox and Anna Franckowiak and E. Friedman and A. Fritz and Philipp Furst and Thomas K. Gaisser and J. Gallagher and E. Ganster and Simone Garrappa and Lisa Marie Gerhardt and A. Ghadimi and Christian Glaser and Theo Glauch and Thorsten Glusenkamp and Azriel Goldschmidt and J. G. Gonzalez and S. Goswami and Darren Grant and T. Gr'egoire and Zachary Griffith and Spencer Griswold and Mustafa Emre Gunduz and Christian Haack and Allan Hallgren and Robert Halliday and L. Halve and Francis Halzen and Martin Ha Minh and Kael D. Hanson and John Hardin and Alexander Harnisch and Andreas Haungs and S. Hauser and Dustin Hebecker and Klaus Helbing and Felix Henningsen and Emma C. Hettinger and Stephanie Virginia Hickford and Joshua Hignight and C. Hill and Gary C. Hill and K. D. Hoffman and Ruth Hoffmann and Tobias Hoinka and Ben Hokanson-Fasig and Kotoyo Hoshina and F. Huang and Martin E. Huber and T. Huber and Klas Hultqvist and Mirco Hunnefeld and Raamis Hussain and Seongjin In and N. Iovine and Aya Ishihara and Mattias Jansson and George S. Japaridze and Minjin Jeong and B. J. P. Jones and R. Joppe and D. Kang and W. Kang and Xiaoping Kang and Alexander Kappes and David Kappesser and Timo Karg and M. Karl and Albrecht Karle and Uli Katz and Matthew Kauer and Moritz Kellermann and John Lawrence Kelley and Ali Kheirandish and J. H. Kim and Ken'ichi Kin and T. Kintscher and Joanna Kiryluk and Spencer R. Klein and Ramesh Koirala and Hermann Kolanoski and Lutz Kopke and Claudio Kopper and Sandro Kopper and D. Jason Koskinen and P. Koundal and M. Kovacevich and Marek Kowalski and Kai Michael Krings and G. Kruckl and Naoko Kurahashi and Andreas L Kyriacou and C. Lagunas Gualda and Justin Lanfranchi and Michael James Larson and F. Lauber and Jeffrey Lazar and K. Leonard and Agnieszka Leszczy'nska and Y. Li and Q. R. Liu and E. Lohfink and C. J. Lozano Mariscal and L. Lu and Fabrizio Lucarelli and A. Ludwig and W. Luszczak and Yang Lyu and W. Y. Ma and J. Madsen and K. Mahn and Yuya Makino and P. Mallik and Sarah Mancina and Ioana Marics and Reina Maruyama and Keiichi Mase and Frank McNally and Kevin J. Meagher and Alberto Martin Gago Medina and M. Meier and Stephan Meighen-Berger and Jennifer Merz and Jessica Micallef and Daniela Mockler and G. Moment'e and Teresa Montaruli and R. W. Moore and Katharina Morik and Robert Morse and M. Moulai and R. Naab and Ryo Nagai and Uwe Naumann and Jannis Necker and L.V.T. Nguyen and Hans Niederhausen and M. U. Nisa and Sarah C. Nowicki and David R. Nygren and A. Obertacke Pollmann and Marie Johanna Oehler and Alexander R. Olivas and E. O’Sullivan and Hershal Pandya and Daria Pankova and N. Park and George K. Parker and E. N. Paudel and Patrick Peiffer and Carlos P'erez de los Heros and Saskia Philippen and Damian Pieloth and Sarah Pieper and Alex Pizzuto and M. Plum and Yuiry Popovych and Alessio Porcelli and M. Prado Rodriguez and P. Buford Price and Brandom Pries and Gerald T. Przybylski and C. Raab and Amirreza Raissi and Mohamed Rameez and Katherine Rawlins and Immacolata Carmen Rea and A. Rehman and Ren{\'e} Reimann and Max Renschler and Giovanni Renzi and Elisa Resconi and Simeon Reusch and Wolfgang Rhode and Michael Richman and Benedikt Riedel and S. M. Robertson and Gerrit Roellinghoff and Martin Rongen and Carsten Rott and Tim Ruhe and Dirk Ryckbosch and D. Rysewyk Cantu and Ibrahim Safa and S. E. Sanchez Herrera and A. Sandrock and J. Sandroos and Marcos Santander and Subir Sarkar and Konstancja Satalecka and M. Scharf and Merlin Schaufel and Harald Schieler and Philipp Schlunder and Thomas Schmidt and A. Schneider and J. Schneider and Frank G. Schroder and L. Schumacher and S. Sclafani and David Seckel and Surujhdeo Seunarine and A. S. Sharma and Shefali Shefali and M. Silva and Barbara Skrzypek and Ben Smithers and Robert Snihur and J. Soedingrekso and Dennis Soldin and G. M. Spiczak and Christian Spiering and J. Stachurska and Michael Stamatikos and Todor Stanev and Robert Stein and J. Stettner and Anna Steuer and Thorsten Stezelberger and Robert G. Stokstad and Timo Sturwald and T. Stuttard and Gregory W. Sullivan and Ignacio J. Taboada and F. Tenholt and Samvel Ter-Antonyan and Serap Tilav and Franziska Tischbein and K. Tollefson and L. Tomankova and Christoph Tonnis and Simona Toscano and Delia Tosi and A. Trettin and Maria Tselengidou and Chun Fai Tung and Andrea Turcati and Roxanne Turcotte and C. F. Turley and Jean Pierre Twagirayezu and Bunheng Ty and M. A. Unland Elorrieta and Nora Valtonen-Mattila and Justin Vandenbroucke and Daan van Eijk and Nick van Eijndhoven and David Vannerom and Jakob van Santen and S. Verpoest and Matthias Vraeghe and Christian Walck and A. Wallace and Travis B. Watson and C. Weaver and Andreas Weindl and Matthew J Weiss and J. Weldert and C. Wendt and Johannes Werthebach and Mark Weyrauch and B. J. Whelan and Nathan Whitehorn and Klaus Wiebe and Christopher Wiebusch and D. R. W. Williams and M. Wolf and Kurt Woschnagg and Gerrit Wrede and Johan Wulff and X. Xu and Y. Xu and Juan Pablo Y{\'a}{\~n}ez and Shigeru Yoshida and T. Yuan and Z. Zhang},
  journal={Journal of Instrumentation},
  year={2021},
  volume={16}
}
Continued improvements on existing reconstruction methods are vital to the success of high-energy physics experiments, such as the IceCube Neutrino Observatory. In IceCube, further challenges arise as the detector is situated at the geographic South Pole where computational resources are limited. However, to perform real-time analyses and to issue alerts to telescopes around the world, powerful and fast reconstruction methods are desired. Deep neural networks can be extremely powerful, and… 

Using convolutional neural networks to reconstruct energy of GeV scale IceCube neutrinos

The IceCube Neutrino Observatory, located under 1.4 km of Antarctic ice, instruments a cubic kilometer of ice with 5,160 optical modules that detect Cherenkov radiation originating from neutrino

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

Separation of track- and shower-like energy deposits in ProtoDUNE-SP using a convolutional neural network

Liquid argon time projection chamber detector technology provides high spatial and calorimetric resolutions on the charged particles traversing liquid argon. As a result, the technology has been used

Graph Neural Networks and Application for Cosmic-Ray Analysis

  • P. Koundal
  • Computer Science
    Proceedings of The 5th International Workshop on Deep Learning in Computational Physics — PoS(DLCP2021)
  • 2021
Graph Neural Networks are outlined and a possible application of using such methods at the IceCube Neutrino Observatory is discussed, reducing the time and computing cost for performing such analysis while boosting sensitivity.

A flexible event reconstruction based on machine learning and likelihood principles

This work shows how the full likelihood for a many-sensor detector can be broken apart into smaller terms, and how neural networks can be trained to approximate all terms solely based on forward simulation, and results in a fast, close-to-optimal surrogate model proportional to the likelihood.

Domain-Informed Neural Networks for Interaction Localization Within Astroparticle Experiments

A domain-informed neural network architecture is proposed that has 60% fewer parameters than an MLP, but that still achieves similar localization performance and provides a path to future architectural developments with improved performance because of their ability to encode additional domain knowledge into the architecture.

Bias and priors in machine learning calibrations for high energy physics

It is argued that the recently proposed Gaussian Ansatz approach can avoid some of the pitfalls of prior dependence, whereas prior-independent data-based calibration remains an open problem.

Low energy event reconstruction in IceCube DeepCore

This work addresses unique challenges associated with the reconstruction of lower energy events in the range of a few to hundreds of GeV and presents two separate, state-of-the-art algorithms for fast directional reconstruction of events based on unscattered light.

Study of mass composition of cosmic rays with IceTop and IceCube

The IceCube Neutrino Observatory is a multi-component detector at the South Pole which detects high-energy particles emerging from astrophysical events. These particles provide us with insights into

References

SHOWING 1-10 OF 54 REFERENCES

Deep Learning in Physics exemplified by the Reconstruction of Muon-Neutrino Events in IceCube

This paper demonstrates how deep learning techniques such as those used in image recognition can be applied to IceCube pulses in order to reconstruct muon-neutrino events.

Event reconstruction for KM3NeT/ORCA using convolutional neural networks

It is shown that this application of deep convolutional neural networks to simulated datasets for a large-volume neutrino telescope yields competitive reconstruction results and performance improvements with respect to classical approaches.

Probing Convolutional Neural Networks for Event Reconstruction in \gamma-Ray Astronomy with Cherenkov Telescopes

Initial results of a CNN-based analysis for background rejection and shower reconstruction are presented, utilizing simulation data from the H.E.S.S., and the influence of image sampling on the performance of the CNN-model predictions are outlined.

Application of Deep Neural Networks to Event Type Classification in IceCube

The classifier that is presented here is based on the modern InceptionResNet architecture and includes multi-task learning in order to broaden the field of application and increase the overall accuracy of the result.

A Convolutional Neural Network Neutrino Event Classifier

This algorithm, CVN (Convolutional Visual Network) identifies neutrino interactions based on their topology without the need for detailed reconstruction and outperforms algorithms currently in use by the NOvA collaboration.

The analysis of VERITAS muon images using convolutional neural networks

Preliminary results of a precise classification of a small set of muon events are presented using a convolutional neural networks model with the raw images as input features, which can be used to calibrate the throughput efficiency of IACTs.

Convolutional neural networks applied to neutrino events in a liquid argon time projection chamber

We present several studies of convolutional neural networks applied to data coming from the MicroBooNE detector, a liquid argon time projection chamber (LArTPC). The algorithms studied include the

Graph Neural Networks for IceCube Signal Classification

This work leverages graph neural networks to improve signal detection in the IceCube neutrino observatory and demonstrates the effectiveness of the GNN architecture on a task classifying IceCube events, where it outperforms both a traditional physics-based method as well as classical 3D convolution neural networks.
...