arXiv subjects

E. A. Huerta

Publications and source records attributed to E. A. Huerta.

At least 55 records · Page 3Linked to original sources

arXiv2020

Deep Learning for Cardiologist-level Myocardial Infarction Detection in Electrocardiograms

Myocardial infarction is the leading cause of death worldwide. In this paper, we design domain-inspired neural network models to detect myocardial infarction. First, we study the contribution of various leads. This systematic analysis, first of its kind in the literature, indicates that out of 15 ECG leads, data from the v6, vz, and ii leads are critical to correctly identify myocardial infarction. Second, we use this finding and adapt the ConvNetQuake neural network model--originally designed to identify earthquakes--to attain state-of-the-art classification results for myocardial infarction, achieving $99.43\%$ classification accuracy on a record-wise split, and $97.83\%$ classification accuracy on a patient-wise split. These two results represent cardiologist-level performance level for myocardial infarction detection after feeding only 10 seconds of raw ECG data into our model. Third, we show that our multi-ECG-channel neural network achieves cardiologist-level performance without the need of any kind of manual feature extraction or data pre-processing.

cs.LG↗

arXiv2020

Artificial neural network subgrid models of 2-D compressible magnetohydrodynamic turbulence

We explore the suitability of deep learning to capture the physics of subgrid-scale ideal magnetohydrodynamics turbulence of 2-D simulations of the magnetized Kelvin-Helmholtz instability. We produce simulations at different resolutions to systematically quantify the performance of neural network models to reproduce the physics of these complex simulations. We compare the performance of our neural networks with gradient models, which are extensively used in the extensively in the magnetohydrodynamic literature. Our findings indicate that neural networks significantly outperform gradient models at reproducing the effects of magnetohydrodynamics turbulence. To the best of our knowledge, this is the first exploratory study on the use of deep learning to learn and reproduce the physics of magnetohydrodynamics turbulence.

physics.comp-ph↗

arXiv2020

Physics-inspired deep learning to characterize the signal manifold of quasi-circular, spinning, non-precessing binary black hole mergers

The spin distribution of binary black hole mergers contains key information concerning the formation channels of these objects, and the astrophysical environments where they form, evolve and coalesce. To quantify the suitability of deep learning to characterize the signal manifold of quasi-circular, spinning, non-precessing binary black hole mergers, we introduce a modified version of WaveNet trained with a novel optimization scheme that incorporates general relativistic constraints of the spin properties of astrophysical black holes. The neural network model is trained, validated and tested with 1.5 million $\ell=|m|=2$ waveforms generated within the regime of validity of NRHybSur3dq8, i.e., mass-ratios $q\leq8$ and individual black hole spins $ | s^z_{\{1,\,2\}} | \leq 0.8$. Using this neural network model, we quantify how accurately we can infer the astrophysical parameters of black hole mergers in the absence of noise. We do this by computing the overlap between waveforms in the testing data set and the corresponding signals whose mass-ratio and individual spins are predicted by our neural network. We find that the convergence of high performance computing and physics-inspired optimization algorithms enable an accurate reconstruction of the mass-ratio and individual spins of binary black hole mergers across the parameter space under consideration. This is a significant step towards an informed utilization of physics-inspired deep learning models to reconstruct the spin distribution of binary black hole mergers in realistic detection scenarios.

gr-qc↗

arXiv2020

Deep Transfer Learning for Star Cluster Classification: I. Application to the PHANGS-HST Survey

We present the results of a proof-of-concept experiment which demonstrates that deep learning can successfully be used for production-scale classification of compact star clusters detected in HST UV-optical imaging of nearby spiral galaxies (D < 20 Mpc) in the PHANGS-HST survey. Given the relatively small nature of existing, human-labelled star cluster samples, we transfer the knowledge of state-of-the-art neural network models for real-object recognition to classify star clusters candidates into four morphological classes. We perform a series of experiments to determine the dependence of classification performance on: neural network architecture (ResNet18 and VGG19-BN); training data sets curated by either a single expert or three astronomers; and the size of the images used for training. We find that the overall classification accuracies are not significantly affected by these choices. The networks are used to classify star cluster candidates in the PHANGS-HST galaxy NGC 1559, which was not included in the training samples. The resulting prediction accuracies are 70%, 40%, 40-50%, 50-70% for class 1, 2, 3 star clusters, and class 4 non-clusters respectively. This performance is competitive with consistency achieved in previously published human and automated quantitative classification of star cluster candidate samples (70-80%, 40-50%, 40-50%, and 60-70%). The methods introduced herein lay the foundations to automate classification for star clusters at scale, and exhibit the need to prepare a standardized dataset of human-labelled star cluster classifications, agreed upon by a full range of experts in the field, to further improve the performance of the networks introduced in this study.

astro-ph.GA↗

arXiv2020

Modeling the uncertainties of solar-system ephemerides for robust gravitational-wave searches with pulsar timing arrays

The regularity of pulsar emissions becomes apparent once we reference the pulses' times of arrivals to the inertial rest frame of the solar system. It follows that errors in the determination of Earth's position with respect to the solar-system barycenter can appear as a time-correlated bias in pulsar-timing residual time series, affecting the searches for low-frequency gravitational waves performed with pulsar timing arrays. Indeed, recent array datasets yield different gravitational-wave background upper limits and detection statistics when analyzed with different solar-system ephemerides. Crucially, the ephemerides do not generally provide usable error representations. In this article we describe the motivation, construction, and application of a physical model of solar-system ephemeris uncertainties, which focuses on the degrees of freedom (Jupiter's orbital elements) most relevant to gravitational-wave searches with pulsar timing arrays. This model, BayesEphem, was used to derive ephemeris-robust results in NANOGrav's 11-yr stochastic-background search, and it provides a foundation for future searches by NANOGrav and other consortia. The analysis and simulations reported here suggest that ephemeris modeling reduces the gravitational-wave sensitivity of the 11-yr dataset; and that this degeneracy will vanish with improved ephemerides and with the longer pulsar timing datasets that will become available in the near future.

astro-ph.HE↗

arXiv2019

The NANOGrav 11-Year Data Set: Limits on Gravitational Wave Memory

The mergers of supermassive black hole binaries (SMBHBs) promise to be incredible sources of gravitational waves (GWs). While the oscillatory part of the merger gravitational waveform will be outside the frequency sensitivity range of pulsar timing arrays (PTAs), the non-oscillatory GW memory effect is detectable. Further, any burst of gravitational waves will produce GW memory, making memory a useful probe of unmodeled exotic sources and new physics. We searched the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) 11-year data set for GW memory. This dataset is sensitive to very low frequency GWs of $\sim3$ to $400$ nHz (periods of $\sim11$ yr $-$ $1$ mon). Finding no evidence for GWs, we placed limits on the strain amplitude of GW memory events during the observation period. We then used the strain upper limits to place limits on the rate of GW memory causing events. At a strain of $2.5\times10^{-14}$, corresponding to the median upper limit as a function of source sky position, we set a limit on the rate of GW memory events at $<0.4$ yr$^{-1}$. That strain corresponds to a SMBHB merger with reduced mass of $ηM \sim 2\times10^{10}M_\odot$ and inclination of $ι=π/3$ at a distance of 1 Gpc. As a test of our analysis, we analyzed the NANOGrav 9-year data set as well. This analysis found an anomolous signal, which does not appear in the 11-year data set. This signal is not a GW, and its origin remains unknown.

astro-ph.HE↗

arXiv2019

Gravitational Wave Denoising of Binary Black Hole Mergers with Deep Learning

Gravitational wave detection requires an in-depth understanding of the physical properties of gravitational wave signals, and the noise from which they are extracted. Understanding the statistical properties of noise is a complex endeavor, particularly in realistic detection scenarios. In this article we demonstrate that deep learning can handle the non-Gaussian and non-stationary nature of gravitational wave data, and showcase its application to denoise the gravitational wave signals generated by the binary black hole mergers GW150914, GW170104, GW170608 and GW170814 from advanced LIGO noise. To exhibit the accuracy of this methodology, we compute the overlap between the time-series signals produced by our denoising algorithm, and the numerical relativity templates that are expected to describe these gravitational wave sources, finding overlaps ${\cal{O}}\gtrsim0.99$. We also show that our deep learning algorithm is capable of removing noise anomalies from numerical relativity signals that we inject in real advanced LIGO data. We discuss the implications of these results for the characterization of gravitational wave signals.

gr-qc↗

arXiv2019

Enabling real-time multi-messenger astrophysics discoveries with deep learning

Multi-messenger astrophysics is a fast-growing, interdisciplinary field that combines data, which vary in volume and speed of data processing, from many different instruments that probe the Universe using different cosmic messengers: electromagnetic waves, cosmic rays, gravitational waves and neutrinos. In this Expert Recommendation, we review the key challenges of real-time observations of gravitational wave sources and their electromagnetic and astroparticle counterparts, and make a number of recommendations to maximize their potential for scientific discovery. These recommendations refer to the design of scalable and computationally efficient machine learning algorithms; the cyber-infrastructure to numerically simulate astrophysical sources, and to process and interpret multi-messenger astrophysics data; the management of gravitational wave detections to trigger real-time alerts for electromagnetic and astroparticle follow-ups; a vision to harness future developments of machine learning and cyber-infrastructure resources to cope with the big-data requirements; and the need to build a community of experts to realize the goals of multi-messenger astrophysics.

gr-qc↗

arXiv2019

The NANOGrav 11-Year Data Set: Evolution of Gravitational Wave Background Statistics

An ensemble of inspiraling supermassive black hole binaries should produce a stochastic background of very low frequency gravitational waves. This stochastic background is predicted to be a power law, with a spectral index of -2/3, and it should be detectable by a network of precisely timed millisecond pulsars, widely distributed on the sky. This paper reports a new "time slicing" analysis of the 11-year data release from the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) using 34 millisecond pulsars. Methods to flag potential "false positive" signatures are developed, including techniques to identify responsible pulsars. Mitigation strategies are then presented. We demonstrate how an incorrect noise model can lead to spurious signals, and show how independently modeling noise across 30 Fourier components, spanning NANOGrav's frequency range, effectively diagnoses and absorbs the excess power in gravitational-wave searches. This results in a nominal, and expected, progression of our gravitational-wave statistics. Additionally we show that the first interstellar medium event in PSR J1713+0747 pollutes the common red noise process with low-spectral index noise, and use a tailored noise model to remove these effects.

astro-ph.HE↗

arXiv2019

Physics of eccentric binary black hole mergers: A numerical relativity perspective

Gravitational wave observations of eccentric binary black hole mergers will provide unequivocal evidence for the formation of these systems through dynamical assembly in dense stellar environments. The study of these astrophysically motivated sources is timely in view of electromagnetic observations, consistent with the existence of stellar mass black holes in the globular cluster M22 and in the Galactic center, and the proven detection capabilities of ground-based gravitational wave detectors. In order to get insights into the physics of these objects in the dynamical, strong-field gravity regime, we present a catalog of 89 numerical relativity waveforms that describe binary systems of non-spinning black holes with mass-ratios $1\leq q \leq 10$, and initial eccentricities as high as $e_0=0.18$ fifteen cycles before merger. We use this catalog to quantify the loss of energy and angular momentum through gravitational radiation, and the astrophysical properties of the black hole remnant, including its final mass and spin, and recoil velocity. We discuss the implications of these results for gravitational wave source modeling, and the design of algorithms to search for and identify eccentric binary black hole mergers in realistic detection scenarios.

gr-qc↗

arXiv2019

Fusing numerical relativity and deep learning to detect higher-order multipole waveforms from eccentric binary black hole mergers

We determine the mass-ratio, eccentricity and binary inclination angles that maximize the contribution of the higher-order waveform multipoles $(\ell, \, |m|)= \{(2,\,2),\, (2,\,1),\, (3,\,3),\, (3,\,2), \, (3,\,1),\, (4,\,4),\, (4,\,3),\, (4,\,2),\,(4,\,1)\}$ for the gravitational wave detection of eccentric binary black hole mergers. We carry out this study using numerical relativity waveforms that describe non-spinning black hole binaries with mass-ratios $1\leq q \leq 10$, and orbital eccentricities as high as $e_0=0.18$ fifteen cycles before merger. For stellar-mass, asymmetric mass-ratio, binary black hole mergers, and assuming LIGO's Zero Detuned High Power configuration, we find that in regions of parameter space where black hole mergers modeled with $\ell=|m|=2$ waveforms have vanishing signal-to-noise ratios, the inclusion of $(\ell, \, |m|)$ modes enables the observation of these sources with signal-to-noise ratios that range between 30\% to 45\% the signal-to-noise ratio of optimally oriented binary black hole mergers modeled with $\ell=|m|=2$ numerical relativity waveforms. Having determined the parameter space where $(\ell, \, |m|)$ modes are important for gravitational wave detection, we construct waveform signals that describe these astrophysically motivate scenarios, and demonstrate that these topologically complex signals can be detected and characterized in real LIGO noise with deep learning algorithms.

gr-qc↗

arXiv2019

Characterization of numerical relativity waveforms of eccentric binary black hole mergers

We introduce a method to quantify the initial eccentricity, gravitational wave frequency, and mean anomaly of numerical relativity simulations that describe non-spinning black holes on moderately eccentric orbits. We demonstrate that this method provides a robust characterization of eccentric binary black hole mergers with mass-ratios $q\leq10$ and eccentricities $e_0\leq0.2$ fifteen cycles before merger. We quantify the circularization rate of a variety of eccentric numerical relativity waveforms introduced in [1] by computing overlaps with their quasi-circular counterparts, finding that $50M$ before merger they attain overlaps ${\cal{O}}\geq0.99$, furnishing evidence for the circularization of moderately eccentric binary black hole mergers with mass-ratios $q\leq10$. We also quantify the importance of including higher-order waveform modes for the characterization of eccentric binary black hole mergers. Using two types of numerical waveforms, one that includes $(\ell, \, |m|)= \{(2,\,2),\, (2,\,1),\, (3,\,3),\, (3,\,2), \, (3,\,1),\, (4,\,4),\, (4,\,3),\, (4,\,2),\,(4,\,1)\}$ and one that only includes the $\ell=|m|=2$ mode, we find that the overlap between these two classes of waveforms is as low as ${\cal{O}}=0.89$ for $q=10$ eccentric binary black hole mergers, underscoring the need to include higher-order waveform modes for the description of these gravitational wave sources. We discuss the implications of these findings for future source modeling and gravitational wave detection efforts.

gr-qc↗

arXiv2019

Deep Learning at Scale for the Construction of Galaxy Catalogs in the Dark Energy Survey

The scale of ongoing and future electromagnetic surveys pose formidable challenges to classify astronomical objects. Pioneering efforts on this front include citizen science campaigns adopted by the Sloan Digital Sky Survey (SDSS). SDSS datasets have been recently used to train neural network models to classify galaxies in the Dark Energy Survey (DES) that overlap the footprint of both surveys. Herein, we demonstrate that knowledge from deep learning algorithms, pre-trained with real-object images, can be transferred to classify galaxies that overlap both SDSS and DES surveys, achieving state-of-the-art accuracy $\gtrsim99.6\%$. We demonstrate that this process can be completed within just eight minutes using distributed training. While this represents a significant step towards the classification of DES galaxies that overlap previous surveys, we need to initiate the characterization of unlabelled DES galaxies in new regions of parameter space. To accelerate this program, we use our neural network classifier to label over ten thousand unlabelled DES galaxies, which do not overlap previous surveys. Furthermore, we use our neural network model as a feature extractor for unsupervised clustering and find that unlabeled DES images can be grouped together in two distinct galaxy classes based on their morphology, which provides a heuristic check that the learning is successfully transferred to the classification of unlabelled DES images. We conclude by showing that these newly labeled datasets can be combined with unsupervised recursive training to create large-scale DES galaxy catalogs in preparation for the Large Synoptic Survey Telescope era.

astro-ph.IM↗

arXiv2019

The NANOGrav 11-Year Data Set: Limits on Gravitational Waves from Individual Supermassive Black Hole Binaries

Observations indicate that nearly all galaxies contain supermassive black holes (SMBHs) at their centers. When galaxies merge, their component black holes form SMBH binaries (SMBHBs), which emit low-frequency gravitational waves (GWs) that can be detected by pulsar timing arrays (PTAs). We have searched the recently-released North American Nanohertz Observatory for Gravitational Waves (NANOGrav) 11-year data set for GWs from individual SMBHBs in circular orbits. As we did not find strong evidence for GWs in our data, we placed 95\% upper limits on the strength of GWs from such sources as a function of GW frequency and sky location. We placed a sky-averaged upper limit on the GW strain of $h_0 < 7.3(3) \times 10^{-15}$ at $f_\mathrm{gw}= 8$ nHz. We also developed a technique to determine the significance of a particular signal in each pulsar using ``dropout' parameters as a way of identifying spurious signals in measurements from individual pulsars. We used our upper limits on the GW strain to place lower limits on the distances to individual SMBHBs. At the most-sensitive sky location, we ruled out SMBHBs emitting GWs with $f_\mathrm{gw}= 8$ nHz within 120 Mpc for $\mathcal{M} = 10^9 \, M_\odot$, and within 5.5 Gpc for $\mathcal{M} = 10^{10} \, M_\odot$. We also determined that there are no SMBHBs with $\mathcal{M} > 1.6 \times 10^9 \, M_\odot$ emitting GWs in the Virgo Cluster. Finally, we estimated the number of potentially detectable sources given our current strain upper limits based on galaxies in Two Micron All-Sky Survey (2MASS) and merger rates from the Illustris cosmological simulation project. Only 34 out of 75,000 realizations of the local Universe contained a detectable source, from which we concluded it was unsurprising that we did not detect any individual sources given our current sensitivity to GWs.

astro-ph.GA↗

arXiv2019

Cyberinfrastructure Requirements to Enhance Multi-messenger Astrophysics

The identification of the electromagnetic counterpart of the gravitational wave event, GW170817, and discovery of neutrinos and gamma-rays from TXS 0506+056 heralded the new era of multi-messenger astrophysics. As the number of multi-messenger events rapidly grow over the next decade, the cyberinfrastructure requirements to handle the increase in data rates, data volume, need for event follow up, and analysis across the different messengers will also explosively grow. The cyberinfrastructure requirements to enhance multi-messenger astrophysics will both be a major challenge and opportunity for astronomers, physicists, computer scientists and cyberinfrastructure specialists. Here we outline some of these requirements and argue for a distributed cyberinfrastructure institute for multi-messenger astrophysics to meet these challenges.

astro-ph.IM↗

arXiv2019

Supporting High-Performance and High-Throughput Computing for Experimental Science

The advent of experimental science facilities-instruments and observatories, such as the Large Hadron Collider, the Laser Interferometer Gravitational Wave Observatory, and the upcoming Large Synoptic Survey Telescope-has brought about challenging, large-scale computational and data processing requirements. Traditionally, the computing infrastructure to support these facility's requirements were organized into separate infrastructure that supported their high-throughput needs and those that supported their high-performance computing needs. We argue that to enable and accelerate scientific discovery at the scale and sophistication that is now needed, this separation between high-performance computing and high-throughput computing must be bridged and an integrated, unified infrastructure provided. In this paper, we discuss several case studies where such infrastructure has been implemented. These case studies span different science domains, software systems, and application requirements as well as levels of sustainability. A further aim of this paper is to provide a basis to determine the common characteristics and requirements of such infrastructure, as well as to begin a discussion of how best to support the computing requirements of existing and future experimental science facilities.

cs.DC↗

arXiv2019

Deep Learning for Multi-Messenger Astrophysics: A Gateway for Discovery in the Big Data Era

This report provides an overview of recent work that harnesses the Big Data Revolution and Large Scale Computing to address grand computational challenges in Multi-Messenger Astrophysics, with a particular emphasis on real-time discovery campaigns. Acknowledging the transdisciplinary nature of Multi-Messenger Astrophysics, this document has been prepared by members of the physics, astronomy, computer science, data science, software and cyberinfrastructure communities who attended the NSF-, DOE- and NVIDIA-funded "Deep Learning for Multi-Messenger Astrophysics: Real-time Discovery at Scale" workshop, hosted at the National Center for Supercomputing Applications, October 17-19, 2018. Highlights of this report include unanimous agreement that it is critical to accelerate the development and deployment of novel, signal-processing algorithms that use the synergy between artificial intelligence (AI) and high performance computing to maximize the potential for scientific discovery with Multi-Messenger Astrophysics. We discuss key aspects to realize this endeavor, namely (i) the design and exploitation of scalable and computationally efficient AI algorithms for Multi-Messenger Astrophysics; (ii) cyberinfrastructure requirements to numerically simulate astrophysical sources, and to process and interpret Multi-Messenger Astrophysics data; (iii) management of gravitational wave detections and triggers to enable electromagnetic and astro-particle follow-ups; (iv) a vision to harness future developments of machine and deep learning and cyberinfrastructure resources to cope with the scale of discovery in the Big Data Era; (v) and the need to build a community that brings domain experts together with data scientists on equal footing to maximize and accelerate discovery in the nascent field of Multi-Messenger Astrophysics.

astro-ph.IM↗

arXiv2018

Probing neutron star structure via f-mode oscillations and damping in dynamical spacetime models

Gravitational wave and electromagnetic observations can provide new insights into the nature of matter at supra-nuclear densities inside neutron stars. Improvements in electromagnetic and gravitational wave sensing instruments continue to enhance the accuracy with which they can measure the masses, radii, and tidal deformability of neutron stars. These better measurements place tighter constraints on the equation of state of cold matter above nuclear density. In this article, we discuss a complementary approach to get insights into the structure of neutron stars by providing a model prediction for non-linear fundamental eigenmodes (f-modes) and their decay over time, which are thought to be induced by time-dependent tides in neutron star binaries. Building on pioneering studies that relate the properties of f-modes to the structure of neutron stars, we systematically study this link in the non-perturbative regime using models that utilize numerical relativity. Using a suite of fully relativistic numerical relativity simulations of oscillating TOV stars, we establish blueprints for the numerical accuracy needed to accurately compute the frequency and damping times of f-mode oscillations, which we expect to be a good guide for the requirements in the binary case. We show that the resulting f-mode frequencies match established results from linear perturbation theory, but the damping times within numerical errors depart from linear predictions. This work lays the foundation for upcoming studies aimed at a comparison of theoretical models of f-mode signatures in gravitational waves, and their uncertainties with actual gravitational wave data, searching for neutron star binaries on highly eccentric orbits, and probing neutron star structure at high densities.

gr-qc↗