Classifying the unknown

Discovering novel gravitational-wave detector glitches using similarity learning

S. Coughlin, S. Bahaadini, N. Rohani, M. Zevin, O. Patane, M. Harandi, C. Jackson, V. Noroozi, S. Allen, J. Areeda, M. Coughlin, P. Ruiz, Christopher Philip Luke Berry, K. Crowston, Aggelos K Katsaggelos, A. Lundgren, C. Østerlund, J. R. Smith, L. Trouille, Vicky Kalogera

Research output: Contribution to journalArticle

Abstract

The observation of gravitational waves from compact binary coalescences by LIGO and Virgo has begun a new era in astronomy. A critical challenge in making detections is determining whether loud transient features in the data are caused by gravitational waves or by instrumental or environmental sources. The citizen-science project Gravity Spy has been demonstrated as an efficient infrastructure for classifying known types of noise transients (glitches) through a combination of data analysis performed by both citizen volunteers and machine learning. We present the next iteration of this project, using similarity indices to empower citizen scientists to create large data sets of unknown transients, which can then be used to facilitate supervised machine-learning characterization. This new evolution aims to alleviate a persistent challenge that plagues both citizen-science and instrumental detector work: the ability to build large samples of relatively rare events. Using two families of transient noise that appeared unexpectedly during LIGO's second observing run, we demonstrate the impact that the similarity indices could have had on finding these new glitch types in the Gravity Spy program.

Original languageEnglish (US)
Article number082002
JournalPhysical Review D
Volume99
Issue number8
DOIs
StatePublished - Apr 15 2019

Fingerprint

LIGO (observatory)
machine learning
classifying
gravitational waves
learning
gravitation
detectors
astronomy
coalescing
iteration

ASJC Scopus subject areas

  • Physics and Astronomy (miscellaneous)

Cite this

Coughlin, S. ; Bahaadini, S. ; Rohani, N. ; Zevin, M. ; Patane, O. ; Harandi, M. ; Jackson, C. ; Noroozi, V. ; Allen, S. ; Areeda, J. ; Coughlin, M. ; Ruiz, P. ; Berry, Christopher Philip Luke ; Crowston, K. ; Katsaggelos, Aggelos K ; Lundgren, A. ; Østerlund, C. ; Smith, J. R. ; Trouille, L. ; Kalogera, Vicky. / Classifying the unknown : Discovering novel gravitational-wave detector glitches using similarity learning. In: Physical Review D. 2019 ; Vol. 99, No. 8.
@article{d89ee250caf842ca81aa45d4c211b1fb,
title = "Classifying the unknown: Discovering novel gravitational-wave detector glitches using similarity learning",
abstract = "The observation of gravitational waves from compact binary coalescences by LIGO and Virgo has begun a new era in astronomy. A critical challenge in making detections is determining whether loud transient features in the data are caused by gravitational waves or by instrumental or environmental sources. The citizen-science project Gravity Spy has been demonstrated as an efficient infrastructure for classifying known types of noise transients (glitches) through a combination of data analysis performed by both citizen volunteers and machine learning. We present the next iteration of this project, using similarity indices to empower citizen scientists to create large data sets of unknown transients, which can then be used to facilitate supervised machine-learning characterization. This new evolution aims to alleviate a persistent challenge that plagues both citizen-science and instrumental detector work: the ability to build large samples of relatively rare events. Using two families of transient noise that appeared unexpectedly during LIGO's second observing run, we demonstrate the impact that the similarity indices could have had on finding these new glitch types in the Gravity Spy program.",
author = "S. Coughlin and S. Bahaadini and N. Rohani and M. Zevin and O. Patane and M. Harandi and C. Jackson and V. Noroozi and S. Allen and J. Areeda and M. Coughlin and P. Ruiz and Berry, {Christopher Philip Luke} and K. Crowston and Katsaggelos, {Aggelos K} and A. Lundgren and C. {\O}sterlund and Smith, {J. R.} and L. Trouille and Vicky Kalogera",
year = "2019",
month = "4",
day = "15",
doi = "10.1103/PhysRevD.99.082002",
language = "English (US)",
volume = "99",
journal = "Physical Review D",
issn = "2470-0010",
publisher = "American Physical Society",
number = "8",

}

Coughlin, S, Bahaadini, S, Rohani, N, Zevin, M, Patane, O, Harandi, M, Jackson, C, Noroozi, V, Allen, S, Areeda, J, Coughlin, M, Ruiz, P, Berry, CPL, Crowston, K, Katsaggelos, AK, Lundgren, A, Østerlund, C, Smith, JR, Trouille, L & Kalogera, V 2019, 'Classifying the unknown: Discovering novel gravitational-wave detector glitches using similarity learning', Physical Review D, vol. 99, no. 8, 082002. https://doi.org/10.1103/PhysRevD.99.082002

Classifying the unknown : Discovering novel gravitational-wave detector glitches using similarity learning. / Coughlin, S.; Bahaadini, S.; Rohani, N.; Zevin, M.; Patane, O.; Harandi, M.; Jackson, C.; Noroozi, V.; Allen, S.; Areeda, J.; Coughlin, M.; Ruiz, P.; Berry, Christopher Philip Luke; Crowston, K.; Katsaggelos, Aggelos K; Lundgren, A.; Østerlund, C.; Smith, J. R.; Trouille, L.; Kalogera, Vicky.

In: Physical Review D, Vol. 99, No. 8, 082002, 15.04.2019.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Classifying the unknown

T2 - Discovering novel gravitational-wave detector glitches using similarity learning

AU - Coughlin, S.

AU - Bahaadini, S.

AU - Rohani, N.

AU - Zevin, M.

AU - Patane, O.

AU - Harandi, M.

AU - Jackson, C.

AU - Noroozi, V.

AU - Allen, S.

AU - Areeda, J.

AU - Coughlin, M.

AU - Ruiz, P.

AU - Berry, Christopher Philip Luke

AU - Crowston, K.

AU - Katsaggelos, Aggelos K

AU - Lundgren, A.

AU - Østerlund, C.

AU - Smith, J. R.

AU - Trouille, L.

AU - Kalogera, Vicky

PY - 2019/4/15

Y1 - 2019/4/15

N2 - The observation of gravitational waves from compact binary coalescences by LIGO and Virgo has begun a new era in astronomy. A critical challenge in making detections is determining whether loud transient features in the data are caused by gravitational waves or by instrumental or environmental sources. The citizen-science project Gravity Spy has been demonstrated as an efficient infrastructure for classifying known types of noise transients (glitches) through a combination of data analysis performed by both citizen volunteers and machine learning. We present the next iteration of this project, using similarity indices to empower citizen scientists to create large data sets of unknown transients, which can then be used to facilitate supervised machine-learning characterization. This new evolution aims to alleviate a persistent challenge that plagues both citizen-science and instrumental detector work: the ability to build large samples of relatively rare events. Using two families of transient noise that appeared unexpectedly during LIGO's second observing run, we demonstrate the impact that the similarity indices could have had on finding these new glitch types in the Gravity Spy program.

AB - The observation of gravitational waves from compact binary coalescences by LIGO and Virgo has begun a new era in astronomy. A critical challenge in making detections is determining whether loud transient features in the data are caused by gravitational waves or by instrumental or environmental sources. The citizen-science project Gravity Spy has been demonstrated as an efficient infrastructure for classifying known types of noise transients (glitches) through a combination of data analysis performed by both citizen volunteers and machine learning. We present the next iteration of this project, using similarity indices to empower citizen scientists to create large data sets of unknown transients, which can then be used to facilitate supervised machine-learning characterization. This new evolution aims to alleviate a persistent challenge that plagues both citizen-science and instrumental detector work: the ability to build large samples of relatively rare events. Using two families of transient noise that appeared unexpectedly during LIGO's second observing run, we demonstrate the impact that the similarity indices could have had on finding these new glitch types in the Gravity Spy program.

UR - http://www.scopus.com/inward/record.url?scp=85065092821&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85065092821&partnerID=8YFLogxK

U2 - 10.1103/PhysRevD.99.082002

DO - 10.1103/PhysRevD.99.082002

M3 - Article

VL - 99

JO - Physical Review D

JF - Physical Review D

SN - 2470-0010

IS - 8

M1 - 082002

ER -