Publications

Gravitational waves from compact binary mergers are now routinely observed by Earth-bound detectors. These observations enable exciting new science, as they have opened a new window to the Universe. However, extracting gravitational-wave signals from the noisy detector data is a challenging problem. The most sensitive search algorithms for compact binary mergers use matched filtering, an algorithm that compares the data with a set of expected template signals. As detectors are upgraded and more sophisticated signal models become available, the number of required templates will increase, which can make some sources computationally prohibitive to search for. The computational cost is of particular concern when low-latency alerts should be issued to maximize the time for electromagnetic follow-up observations. One potential solution to reduce computational requirements that has started to be explored in the last decade is machine learning. However, different proposed deep learning searches target varying parameter spaces and use metrics that are not always comparable to existing literature. Consequently, a clear picture of the capabilities of machine learning searches has been sorely missing. In this thesis, we closely examine the sensitivity of various deep learning gravitational-wave search algorithms and introduce new methods to detect signals from binary black hole and binary neutron star mergers at previously untested statistical confidence levels. By using the sensitive distance as our core metric, we allow for a direct comparison of our algorithms to state-of-the-art search pipelines. As part of this thesis, we organized a global mock data challenge to create a benchmark for machine learning search algorithms targeting compact binaries. This way, the tools developed in this thesis are made available to the greater community by publishing them as open source software. Our studies show that, depending on the parameter space, deep learning gravitational-wave search algorithms are already competitive with current production search pipelines. We also find that strategies developed for traditional searches can be effectively adapted to their machine learning counterparts. In regions where matched filtering becomes computationally expensive, available deep learning algorithms are also limited in their capability. We find reduced sensitivity to long duration signals compared to the excellent results for short-duration binary black hole signals.

We present the results of the first Machine Learning Gravitational-Wave Search Mock Data Challenge (MLGWSC-1). 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 seconds with the inclusion of precession effects and higher order modes. We present the average sensitivity distance and runtime 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.

Gravitational waves from the coalescence of compact-binary sources are now routinely observed by Earth bound detectors. The most sensitive search algorithms convolve many different pre-calculated gravitational waveforms with the detector data and look for coincident matches between different detectors. Machine learning is being explored as an alternative approach to building a search algorithm that has the prospect to reduce computational costs and target more complex signals. In this work we construct a two-detector search for gravitational waves from binary black hole mergers using neural networks trained on non-spinning binary black hole data from a single detector. The network is applied to the data from both observatories independently and we check for events coincident in time between the two. This enables the efficient analysis of large quantities of background data by time-shifting the independent detector data. We find that while for a single detector the network retains 91.5% of the sensitivity matched filtering can achieve, this number drops to 83.9% for two observatories. To enable the network to check for signal consistency in the detectors, we then construct a set of simple networks that operate directly on data from both detectors. We find that none of these simple two-detector networks are capable of improving the sensitivity over applying networks individually to the data from the detectors and searching for time coincidences.

Compact binary systems emit gravitational radiation which is potentially detectable by current Earth bound detectors. Extracting these signals from the instruments' background noise is a complex problem and the computational cost of most current searches depends on the complexity of the source model. Deep learning may be capable of finding signals where current algorithms hit computational limits. Here we restrict our analysis to signals from non-spinning binary black holes and systematically test different strategies by which training data is presented to the networks. To assess the impact of the training strategies, we re-analyze the first published networks and directly compare them to an equivalent matched-filter search. We find that the deep learning algorithms can generalize low signal-to-noise ratio (SNR) signals to high SNR ones but not vice versa. As such, it is not beneficial to provide high SNR signals during training, and fastest convergence is achieved when low SNR samples are provided early on. During testing we found that the networks are sometimes unable to recover any signals when a false alarm probability <10^-3 is required. We resolve this restriction by applying a modification we call unbounded Softmax replacement (USR) after training. With this alteration we find that the machine learning search retains ≥97.5% of the sensitivity of the matched-filter search down to a false-alarm rate of 1 per month.

We present the third open gravitational-wave catalog (3-OGC) of compact-binary coalescences, based on the analysis of the public LIGO and Virgo data from 2015 through 2019 (O1, O2, O3a). Our updated catalog includes a population of 57 observations, including 4 binary black hole mergers that had not been previously reported. This consists of 55 binary black hole mergers and the 2 binary neutron star mergers, GW170817 and GW190425. We find no additional significant binary neutron star or neutron star–black hole merger events. The most confident new detection is the binary black hole merger GW190925_232845, which was observed by the LIGO–Hanford and Virgo observatories with its primary and secondary component masses are and , respectively. We estimate the parameters of all binary black hole events using an up-to-date waveform model that includes both subdominant harmonics and precession effects. To enable deep follow up as our understanding of the underlying populations evolves, we make available our comprehensive catalog of events, including the subthreshold population of candidates, and the posterior samples of our source parameter estimates.

We present the fourth Open Gravitational-wave Catalog (4-OGC) of binary neutron star (BNS), binary black hole (BBH) and neutron star-black hole (NSBH) mergers. The catalog includes observations from 2015-2020 covering the first through third observing runs (O1, O2, O3a, O3b) of Advanced LIGO and Advanced Virgo. The updated catalog includes 7 BBH mergers which were not previously reported with high significance during O3b for a total of 94 observations: 90 BBHs, 2 NSBHs, and 2 BNSs. The most confident new detection, GW200318_191337, has component masses 49.1 (−12.0, +16.4) Msun and 31.6 (−11.6, +12.0) Msun​; its redshift of 0.84 (−0.35, +0.4) (90 credible interval) may make it the most distant merger so far. We provide reference parameter estimates for each of these sources using an up-to-date model accounting for instrumental calibration uncertainty. The corresponding data release also includes our full set of sub-threshold candidates.

We present the prospects for the early (pre-merger) detection and localization of compact-binary coalescences using gravitational waves over the next 10 yr. Early warning can enable the direct observation of the prompt and early electromagnetic emission of a neutron star merger. We examine the capabilities of the ground-based detectors at their “Design” sensitivity (2021–2022), the planned “A+” upgrade (2024–2026), and the envisioned “Voyager” concept (late 2020s). We find that for a fiducial rate of binary neutron star mergers of 1000 Gpc^−3 yr^−1, the Design, A+, and Voyager era networks can provide 18, 54, and 195 s of warning for one source per year of observing, respectively, with a sky localization area <100 deg^2 at a 90% credible level. At the same rate, the A+ and Voyager era networks will be able to provide 9 and 43 s of warning, respectively, for a source with <10 deg^2 localization area. We compare the idealized search sensitivity to that achieved by the PyCBC Live search tuned for pre-merger detection. The gravitational-wave community will be prepared to produce pre-merger alerts. Our results motivate the operation of observatories with wide fields of view, automation, and the capability for fast slewing to observe simultaneously with the gravitational-wave network.

As two neutron stars merge, they emit gravitational waves that can potentially be detected by Earth-bound detectors. Matched-filtering-based algorithms have traditionally been used to extract quiet signals embedded in noise. We introduce a novel neural-network-based machine learning algorithm that uses time series strain data from gravitational-wave detectors to detect signals from nonspinning binary neutron star mergers. For the Advanced LIGO design sensitivity, our network has an average sensitive distance of 130 Mpc at a false-alarm rate of ten per month. Compared to other state-of-the-art machine learning algorithms, we find an improvement by a factor of 4 in sensitivity to signals with a signal-to-noise ratio between 8 and 15. However, this approach is not yet competitive with traditional matched-filtering-based methods. A conservative estimate indicates that our algorithm introduces on average 10.2 s of latency between signal arrival and generating an alert. We give an exact description of our testing procedure, which can be applied not only to machine-learning-based algorithms but all other search algorithms as well. We thereby improve the ability to compare machine learning and classical searches.

Gravitational waves are now observed routinely. Therefore, data analysis has to keep up with ever improving detectors. One relatively new tool to search for gravitational wave signals in detector data are machine learning algorithms that utilize deep neuralnnetworks. The first successful application was able to differentiate time series strain data that contains a gravitational wave from a binary black hole merger from data that consists purely of noise. This work expands the analysis to signals from binary neutron star mergers, where a rapid detection is most valuable, as electromagnetic counterparts might otherwise be missed or not observed for long enough. We showcase many different architecture, discuss what choices improved the sensitivity of our search and introduce a new multi-rate approach. We find that the final algorithm gives state of the art performance in comparison to other search pipelines that use deep neural networks. On the other hand we also conclude that our analysis is not yet able to achieve sensitivities that are on par with template based searches. We report our results at false alarm rates down to ~30 samples/month which has not been tested by other neural network algorithms. We hope to provide information about useful architectural choices and improve our algorithm in the future to achieve sensitivities and false alarm rates that rival matched filtering based approaches on.