An official website of the United States government
The first successful detection of gravitational waves by ground-based observatories, such as the Laser Interferometer Gravitational-Wave Observatory (LIGO), marked a breakthrough in our comprehension of the Universe. However, due to the unprecedented sensitivity required to make such observations, gravitational-wave detectors also capture disruptive noise sources called glitches, which can potentially be confused for or mask gravitational-wave signals. To address this problem, a community-science project, Gravity Spy, incorporates human insight and machine learning to classify glitches in LIGO data. The machine-learning classifier, integrated into the project since 2017, has evolved over time to accommodate increasing numbers of glitch classes. Despite its success, limitations have arisen in the ongoing LIGO fourth observing run (O4) due to the architecture's simplicity, which led to poor generalization and inability to handle multi-time window inputs effectively. We propose an advanced classifier for O4 glitches. Using data from previous observing runs, we evaluate different fusion strategies for multi-time window inputs, using label smoothing to counter noisy labels, and enhancing interpretability through attention module-generated weights. Our new O4 classifier shows improved performance, and will enhance glitch classification, aiding in the ongoing exploration of gravitational-wave phenomena.
more »
« less
Advancing Glitch Classification in Gravity Spy: Multi-view Fusion with Attention-based Machine Learning for Advanced LIGO’s Fourth Observing Run
The first successful detection of gravitational waves by ground-based observatories, such as the Laser Interferometer Gravitational-Wave Observatory (LIGO), marked a revolutionary breakthrough in our comprehension of the Universe. However, due to the unprecedented sensitivity required to make such observations, gravitational-wave detectors also capture disruptive noise sources called glitches, potentially masking or appearing as gravitational wave signals themselves. To address this problem, a community-science project, Gravity Spy, incorporates human insight and machine learning to classify glitches in LIGO data. The machine learning classifier, integrated into the project since 2017, has evolved over time to accommodate increasing numbers of glitch classes. Despite its success, limitations have arisen in the ongoing LIGO fourth observing run (O4) due to its architecture’s simplicity, which led to poor generalization and inability to handle multi-time window inputs effectively. We propose an advanced classifier for O4 glitches. Our contributions include evaluating fusion strategies for multi-time window inputs, using label smoothing to counter noisy labels, and enhancing interpretability through attention module-generated weights. This development seeks to enhance glitch classification, aiding in the ongoing exploration of gravitational-wave phenomena.
more »
« less
- Award ID(s):
- 2107334
- PAR ID:
- 10544840
- Publisher / Repository:
- Arxivorg
- Date Published:
- Journal Name:
- arXivorg
- ISSN:
- 2331-8422
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
In this paper we investigate the impact of transient noise artifacts, or glitches, on gravitational- wave inference from ground-based interferometer data, and test how modeling and subtracting these glitches affects the inferred parameters. Due to their time-frequency morphology, broadband glitches cause moderate to significant biasing of posterior distributions away from true values. In contrast, narrowband glitches induce negligible biasing effects, due to distinct signal and glitch morphologies. We inject simulated binary black hole signals into data containing three occurring glitch types from past LIGO-Virgo observing runs, and reconstruct both signal and glitch waveforms using BayesWave, a wavelet-based Bayesian analysis. We apply the standard LIGO-Virgo-KAGRA deglitching pro- cedure to the detector data, which consists of subtracting from calibrated LIGO data the glitch waveform estimated by the joint BayesWave inference. We produce posterior distributions on the parameters of the injected signal before and after subtracting the glitch, and we show that removing the transient noise effectively mitigates bias from broadband glitches. This study provides a baseline validation of existing techniques, while demonstrating waveform reconstruction improvements to the Bayesian algorithm for robust astrophysical characterization in glitch-prone detector data.more » « less
-
Understanding the noise in gravitational-wave detectors is central to detecting and interpreting gravitational-wave signals. Glitches are transient, non-Gaussian noise features that can have a range of environmental and instrumental origins. The Gravity Spy project uses a machine-learning algorithm to classify glitches based upon their time–frequency morphology. The resulting set of classified glitches can be used as input to detector-characterisation investigations of how to mitigate glitches, or data-analysis studies of how to ameliorate the impact of glitches. Here we present the results of the Gravity Spy analysis of data up to the end of the third observing run of Advanced LIGO. We classify 233981 glitches from LIGO Hanford and 379805 glitches from LIGO Livingston into morphological classes. We find that the distribution of glitches differs between the two LIGO sites. This highlights the potential need for studies of data quality to be individually tailored to each gravitational-wave observatory.more » « less
-
Gravitational waves from black hole binary mergers carry information about the component spins, but inference is sensitive to analysis assumptions, which may be broken by terrestrial noise transients known as glitches. Using a variety of simulated glitches and gravitational wave signals, we study the conditions under which glitches can bias spin measurements. We confirm the theoretical expectation that inference and subtraction of glitches invariably leaves behind residual power due to statistical uncertainty, no matter the strength (signal-to-noise ratio, SNR) of the original glitch. Next we show that low-SNR glitches — including those below the threshold for flagging data-quality issues — can still significantly bias spin inference. Such biases occur for a range of glitch morphologies, even in cases where glitches and signals are not precisely aligned in phase. Furthermore, we find that residuals of glitch subtraction can result in biases as well. Our results suggest that joint inference of the glitch and gravitational wave parameters, with appropriate models and priors, is required to address these uncertainties inherent in glitch mitigation via subtraction.more » « less
-
Gravity Spy is a citizen science project that draws on the contributions of both humans and machines to achieve its scientific goals. The system supports the Laser Interferometer Gravitational Observatory (LIGO) by classifying “glitches” that interfere with observations. The system makes three advances on the current state of the art: explicit training for new volunteers, synergy between machine and human classification and support for discovery of new classes of glitch. As well, it provides a platform for human-centred computing research on motivation, learning and collaboration. The system has been launched and is currently in operation.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript
- Journal Article:
- The DOI is not currently available.
