Enhancing gravitational-wave science with machine learning

Accelerated Design for Perovskite-Oxide-Based Photocatalysts Using Machine Learning Techniques

The rapid discovery of photocatalysts with desired performance among tens of thousands of potential perovskites represents a significant advancement. To expedite the design of perovskite-oxide-based photocatalysts, we developed a model of ABO3-type perovskites using machine learning methods based on atomic and experimental parameters. This model can be used to predict specific surface area (SSA), a key parameter closely associated with photocatalytic activity. The model construction involved several steps, including data collection, feature selection, model construction, web-service development, virtual screening and mechanism elucidation. Statistical analysis revealed that the support vector regression model achieved a correlation coefficient of 0.9462 for the training set and 0.8786 for the leave-one-out cross-validation. The potential perovskites with higher SSA than the highest SSA observed in the existing dataset were identified using the model and our computation platform. We also developed a webserver of the model, freely accessible to users. The methodologies outlined in this study not only facilitate the discovery of new perovskites but also enable exploration of the correlations between the perovskite properties and the physicochemical features. These findings provide valuable insights for further research and applications of perovskites using machine learning techniques.

Normalizing Flows as an Avenue to Studying Overlapping Gravitational Wave Signals

Because of its speed after training, machine learning is often envisaged as a solution to a manifold of the issues faced in gravitational-wave astronomy. Demonstrations have been given for various applications in gravitational-wave data analysis. In this Letter, we focus on a challenging problem faced by third-generation detectors: parameter inference for overlapping signals. Because of the high detection rate and increased duration of the signals, they will start to overlap, possibly making traditional parameter inference techniques difficult to use. Here, we show a proof-of-concept application of normalizing flows to perform parameter estimation on overlapped binary black hole systems.

Neural Importance Sampling for Rapid and Reliable Gravitational-Wave Inference

We combine amortized neural posterior estimation with importance sampling for fast and accurate gravitational-wave inference. We first generate a rapid proposal for the Bayesian posterior using neural networks, and then attach importance weights based on the underlying likelihood and prior. This provides (1) a corrected posterior free from network inaccuracies, (2) a performance diagnostic (the sample efficiency) for assessing the proposal and identifying failure cases, and (3) an unbiased estimate of the Bayesian evidence. By establishing this independent verification and correction mechanism we address some of the most frequent criticisms against deep learning for scientific inference. We carry out a large study analyzing 42 binary black hole mergers observed by LIGO and Virgo with the SEOBNRv4PHM and IMRPhenomXPHM waveform models. This shows a median sample efficiency of ≈10% (2 orders of magnitude better than standard samplers) as well as a tenfold reduction in the statistical uncertainty in the log evidence. Given these advantages, we expect a significant impact on gravitational-wave inference, and for this approach to serve as a paradigm for harnessing deep learning methods in scientific applications.

Gravitational Wave-Signal Recognition Model Based on Fourier Transform and Convolutional Neural Network

The recent detection of gravitational waves is a remarkable milestone in the history of astrophysics. With the further development of gravitational wave detection technology, traditional filter-matching methods no longer meet the needs of signal recognition. Thus, it is imperative that we develop new methods. In this study, we apply a gravitational wave signal recognition model based on Fourier transformation and a convolutional neural network (CNN). The gravitational wave time-domain signal is transformed into a 2D frequency-domain signal graph for feature recognition using a CNN model. Experimental results reveal that the frequency-domain signal graph provides a better feature description of the gravitational wave signal than that provided by the time-domain signal. Our method takes advantage of the CNN's convolution computation to improve the accuracy of signal recognition. The impact of the training set size and image filtering on the performance of the developed model is also evaluated. Additionally, the Resnet101 model, developed on the Baidu EasyDL platform, is adopted as a comparative model. Our average recognition accuracy performs approximately 4% better than the Resnet101 model. Based on the excellent performance of convolutional neural network in the field of image recognition, this paper studies the characteristics of gravitational wave signals and obtains a more appropriate recognition model after training and tuning, in order to achieve the purpose of automatic recognition of whether the signal data contain real gravitational wave signals.

Outbound Data Legality Analysis in CPTPP Countries under the Environment of Cross-Border Data Flow Governance

The governance of cross-border data flows around the digital economy, data security, and data sovereignty has become a crucial global governance issue. This paper evaluates the legitimacy of data exit rules of CPTPP countries based on machine learning algorithm models under the perspective of cross-border data flow governance. In this study, four machine learning algorithms, namely, logistic regression, decision tree, random forest, and GBDT, are used to build an outbound data assessment and evaluation model. The confusion matrix is used to classify the outbound data legitimacy dichotomously. The recall, precision, and F1 scores are evaluated to compare the empirical results of each model. Based on this, a logistic regression-based outbound data risk scoring model is introduced to quantify the outbound data risk at a deeper level and to classify the outbound data risk level for the reference of regulators to make more scientific and reasonable decisions. The experimental results show that the machine learning models can meet the needs and applications of practical work and make accurate predictions of outbound data risks.

Computational challenges for multimodal astrophysics

In the coming decades, we will face major computational challenges, when the improved sensitivity of third-generation gravitational wave detectors will be such that they will be able to detect a high number (of the order of 7 × 104 per year) of multi-messenger events from binary neutron star mergers, similar to GW 170817. In this Perspective, we discuss the application of multimodal artificial intelligence techniques for multi-messenger astrophysics, fusing the information from different signal emissions.

Applications and Techniques for Fast Machine Learning in Science

In this community review report, we discuss applications and techniques for fast machine learning (ML) in science-the concept of integrating powerful ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs.

Nonlinear Noise Cleaning in Gravitational-Wave Detectors With Convolutional Neural Networks

Currently, the sub-60 Hz sensitivity of gravitational-wave (GW) detectors like Advanced LIGO (aLIGO) is limited by the control noises from auxiliary degrees of freedom which nonlinearly couple to the main GW readout. One promising way to tackle this challenge is to perform nonlinear noise mitigation using convolutional neural networks (CNNs), which we examine in detail in this study. In many cases, the noise coupling is bilinear and can be viewed as a few fast channels' outputs modulated by some slow channels. We show that we can utilize this knowledge of the physical system and adopt an explicit “slow×fast” structure in the design of the CNN to enhance its performance of noise subtraction. We then examine the requirements in the signal-to-noise ratio (SNR) in both the target channel (i.e., the main GW readout) and in the auxiliary sensors in order to reduce the noise by at least a factor of a few. In the case of limited SNR in the target channel, we further demonstrate that the CNN can still reach a good performance if we use curriculum learning techniques, which in reality can be achieved by combining data from quiet times and those from periods with active noise injections.

Inference-Optimized AI and High Performance Computing for Gravitational Wave Detection at Scale

We introduce an ensemble of artificial intelligence models for gravitational wave detection that we trained in the Summit supercomputer using 32 nodes, equivalent to 192 NVIDIA V100 GPUs, within 2 h. Once fully trained, we optimized these models for accelerated inference using NVIDIA TensorRT. We deployed our inference-optimized AI ensemble in the ThetaGPU supercomputer at Argonne Leadership Computer Facility to conduct distributed inference. Using the entire ThetaGPU supercomputer, consisting of 20 nodes each of which has 8 NVIDIA A100 Tensor Core GPUs and 2 AMD Rome CPUs, our NVIDIA TensorRT-optimized AI ensemble processed an entire month of advanced LIGO data (including Hanford and Livingston data streams) within 50 s. Our inference-optimized AI ensemble retains the same sensitivity of traditional AI models, namely, it identifies all known binary black hole mergers previously identified in this advanced LIGO dataset and reports no misclassifications, while also providing a 3X inference speedup compared to traditional artificial intelligence models. We used time slides to quantify the performance of our AI ensemble to process up to 5 years worth of advanced LIGO data. In this synthetically enhanced dataset, our AI ensemble reports an average of one misclassification for every month of searched advanced LIGO data. We also present the receiver operating characteristic curve of our AI ensemble using this 5 year long advanced LIGO dataset. This approach provides the required tools to conduct accelerated, AI-driven gravitational wave detection at scale.

Detector Characterization and Mitigation of Noise in Ground-Based Gravitational-Wave Interferometers

Since the early stages of operation of ground-based gravitational-wave interferometers, careful monitoring of these detectors has been an important component of their successful operation and observations. Characterization of gravitational-wave detectors blends computational and instrumental methods of investigating the detector performance. These efforts focus both on identifying ways to improve detector sensitivity for future observations and understand the non-idealized features in data that has already been recorded. Alongside a focus on the detectors themselves, detector characterization includes careful studies of how astrophysical analyses are affected by different data quality issues. This article presents an overview of the multifaceted aspects of the characterization of interferometric gravitational-wave detectors, including investigations of instrumental performance, characterization of interferometer data quality, and the identification and mitigation of data quality issues that impact analysis of gravitational-wave events. Looking forward, we discuss efforts to adapt current detector characterization methods to meet the changing needs of gravitational-wave astronomy.

Real-Time Gravitational Wave Science with Neural Posterior Estimation

We demonstrate unprecedented accuracy for rapid gravitational wave parameter estimation with deep learning. Using neural networks as surrogates for Bayesian posterior distributions, we analyze eight gravitational wave events from the first LIGO-Virgo Gravitational-Wave Transient Catalog and find very close quantitative agreement with standard inference codes, but with inference times reduced from O(day) to 20 s per event. Our networks are trained using simulated data, including an estimate of the detector noise characteristics near the event. This encodes the signal and noise models within millions of neural-network parameters and enables inference for any observed data consistent with the training distribution, accounting for noise nonstationarity from event to event. Our algorithm-called "DINGO"-sets a new standard in fast and accurate inference of physical parameters of detected gravitational wave events, which should enable real-time data analysis without sacrificing accuracy.

Search Methods for Continuous Gravitational-Wave Signals from Unknown Sources in the Advanced-Detector Era

Continuous gravitational waves are long-lasting forms of gravitational radiation produced by persistent quadrupolar variations of matter. Standard expected sources for ground-based interferometric detectors are neutron stars presenting non-axisymmetries such as crustal deformations, r-modes or free precession. More exotic sources could include decaying ultralight boson clouds around spinning black holes. A rich suite of data-analysis methods spanning a wide bracket of thresholds between sensitivity and computational efficiency has been developed during the last decades to search for these signals. In this work, we review the current state of searches for continuous gravitational waves using ground-based interferometer data, focusing on searches for unknown sources. These searches typically consist of a main stage followed by several post-processing steps to rule out outliers produced by detector noise. So far, no continuous gravitational wave signal has been confidently detected, although tighter upper limits are placed as detectors and search methods are further developed.

Statistically-informed deep learning for gravitational wave parameter estimation

We introduce deep learning models to estimate the masses of the binary components of black hole mergers, ( m 1 , m 2 ) , and three astrophysical properties of the post-merger compact remnant, namely, the final spin, a f , and the frequency and damping time of the ringdown oscillations of the fundamental ℓ = m = 2 bar mode, ( ω R , ω I ) . Our neural networks combine a modified WaveNet architecture with contrastive learning and normalizing flow. We validate these models against a Gaussian conjugate prior family whose posterior distribution is described by a closed analytical expression. Upon confirming that our models produce statistically consistent results, we used them to estimate the astrophysical parameters ( m 1 , m 2 , a f , ω R , ω I ) of five binary black holes: GW150914, GW170104, GW170814, GW190521 and GW190630. We use PyCBC Inference to directly compare traditional Bayesian methodologies for parameter estimation with our deep learning based posterior distributions. Our results show that our neural network models predict posterior distributions that encode physical correlations, and that our data-driven median results and 90% confidence intervals are similar to those produced with gravitational wave Bayesian analyses. This methodology requires a single V100 NVIDIA GPU to produce median values and posterior distributions within two milliseconds for each event. This neural network, and a tutorial for its use, are available at the Data and Learning Hub for Science.

NNETFIX: an artificial neural network-based denoising engine for gravitational-wave signals

Instrumental and environmental transient noise bursts in gravitational-wave (GW) detectors, or glitches, may impair astrophysical observations by adversely affecting the sky localization and the parameter estimation of GW signals. Denoising of detector data is especially relevant during low-latency operations because electromagnetic follow-up of candidate detections requires accurate, rapid sky localization and inference of astrophysical sources. NNETFIX is a machine learning, artificial neural network-based algorithm designed to estimate the data containing a transient GW signal with an overlapping glitch as though the glitch was absent. The sky localization calculated from the denoised data may be significantly more accurate than the sky localization obtained from the original data or by removing the portion of the data impacted by the glitch. We test NNETFIX in simulated scenarios of binary black hole coalescence signals and discuss the potential for its use in future low-latency LIGO-Virgo-KAGRA searches. In the majority of cases for signals with a high signal-to-noise ratio, we find that the overlap of the sky maps obtained with the denoised data and the original data is better than the overlap of the sky maps obtained with the original data and the data with the glitch removed.

Complete parameter inference for GW150914 using deep learning

The LIGO and Virgo gravitational-wave observatories have detected many exciting events over the past 5 years. To infer the system parameters, iterative sampling algorithms such as MCMC are typically used with Bayes’ theorem to obtain posterior samples—by repeatedly generating waveforms and comparing to measured strain data. However, as the rate of detections grows with detector sensitivity, this poses a growing computational challenge. To confront this challenge, as well as that of fast multimessenger alerts, in this study we apply deep learning to learn non-iterative surrogate models for the Bayesian posterior. We train a neural-network conditional density estimator to model posterior probability distributions over the full 15-dimensional space of binary black hole system parameters, given detector strain data from multiple detectors. We use the method of normalizing flows—specifically, a neural spline flow—which allows for rapid sampling and density estimation. Training the network is likelihood-free, requiring samples from the data generative process, but no likelihood evaluations. Through training, the network learns a global set of posteriors: it can generate thousands of independent posterior samples per second for any strain data consistent with the training distribution. We demonstrate our method by performing inference on GW150914, and obtain results in close agreement with standard techniques.