Physical Sciences Division
Published January 27, 2023 | Version v1
Journal article Open

First machine learning gravitational-wave search mock data challenge

Creators

  • 1. Max-Planck-Institut für Gravitationsphysik
  • 2. Friedrich-Schiller-Universität Jena
  • 3. Chinese Academy of Sciences
  • 4. Beijing Normal University
  • 5. Peng Cheng Laboratory
  • 6. Aristotle University of Thessaloniki
  • 7. Università di Trento
  • 8. INFN
  • 9. University of Florida
  • 10. European Gravitational Observatory
  • 11. University of Chicago
  • 12. University of Glasgow

Description

We present the results of the first Machine Learning Gravitational-Wave Search Mock Data Challenge. For this challenge, participating groups had to identify gravitational-wave signals from binary black hole mergers of increasing complexity and duration embedded in progressively more realistic noise. The final of the 4 provided datasets contained real noise from the O3a observing run and signals up to a duration of 20 s with the inclusion of precession effects and higher order modes. We present the average sensitivity distance and run-time for the 6 entered algorithms derived from 1 month of test data unknown to the participants prior to submission. Of these, 4 are machine learning algorithms. We find that the best machine learning based algorithms are able to achieve up to 95% of the sensitive distance of matched-filtering based production analyses for simulated Gaussian noise at a false-alarm rate (FAR) of one per month. In contrast, for real noise, the leading machine learning search achieved 70%. For higher FARs the differences in sensitive distance shrink to the point where select machine learning submissions outperform traditional search algorithms at FARs $≥200$ per month on some datasets. Our results show that current machine learning search algorithms may already be sensitive enough in limited parameter regions to be useful for some production settings. To improve the state-of-the-art, machine learning algorithms need to reduce the false-alarm rates at which they are capable of detecting signals and extend their validity to regions of parameter space where modeled searches are computationally expensive to run. Based on our findings we compile a list of research areas that we believe are the most important to elevate machine learning searches to an invaluable tool in gravitational-wave signal detection.

Files

PhysRevD.107.023021.pdf

Files (4.3 MB)

Name Size Download all
md5:31d097f4ed37b390455011b91fe3f1d9
4.3 MB Preview Download

Additional details

Identifiers

DOI
10.1103/PhysRevD.107.023021
Other
oai:uchicago.tind.io:13055

Funding

Peng Cheng Laboratory
Cloud Brain
National Key Research and Development Program of China
2021YFC2203001
NSFC
11920101003
NSFC
12021003
CAS Project for Young Scientists in Basic Research
YSBR-006
Aristotle University of Thessaloniki
IT Center
European Research Council
European Union’s Horizon 2020 research and innovation programme
National Science Foundation
PHY 1806165
National Science Foundation
PHY 2110060
National Science Foundation
OAC-2209892
National Science Foundation
OAC-1931561
Science and Technology Research Council
ST/V005634/1
European Cooperation in Science and Technology
CA17137
Max Planck Society
Independent Research Group Programme
European Gravitational Observatory
French Centre National de Recherche Scientifique
Italian Istituto Nazionale di Fisica Nucleare
Dutch Nikhef
Ministry of Education, Culture, Sports, Science and Technology
Japan Society for the Promotion of Science
National Research Foundation
Ministry of Science and ICT
Academia Sinica
Ministry of Science and Technology

UChicago Information

Division(s)
Physical Sciences Division
Department(s)
Computer Science