Comparison of neural network architectures for feature extraction from binary black hole merger waveforms
arXiv:2307.16668 · doi:10.1088/2632-2153/ad2972
Abstract
We evaluate several neural-network architectures, both convolutional and recurrent, for gravitational-wave time-series feature extraction by performing point parameter estimation on noisy waveforms from binary-black-hole mergers. We build datasets of 100,000 elements for each of four different waveform models (or approximants) in order to test how approximant choice affects feature extraction. Our choices include \texttt{SEOBNRv4P} and \texttt{IMRPhenomPv3}, which contain only the dominant quadrupole emission mode, alongside \texttt{IMRPhenomPv3HM} and \texttt{NRHybSur3dq8}, which also account for high-order modes. Each dataset element is injected into detector noise corresponding to the third observing run of the LIGO-Virgo-KAGRA (LVK) collaboration. We identify the Temporal Convolutional Network (TCN) architecture as the overall best performer in terms of training and validation losses and absence of overfitting to data. Comparison of results between datasets shows that the choice of waveform approximant for the creation of a dataset conditions the feature extraction ability of a trained network. Hence, care should be taken when building a dataset for the training of neural networks, as certain approximants may result in better network convergence of evaluation metrics. However, this performance does not necessarily translate to data which is more faithful to numerical relativity simulations. We also apply this network on actual signals from LVK runs, finding that its feature-extracting performance can be effective on real data.
arXiv commentFixed Fig. 7 not being rendered
References in corpus (47)
- Adam: A Method for Stochastic Optimization
- Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
- GW170817: Observation of Gravitational Waves from a Binary Neutron Star Inspiral
- An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling
- Advanced Virgo: a 2nd generation interferometric gravitational wave detector
- GWTC-1: A Gravitational-Wave Transient Catalog of Compact Binary Mergers Observed by LIGO and Virgo during the First and Second Observing Runs
- Advanced LIGO
- GWTC-2: Compact Binary Coalescences Observed by LIGO and Virgo During the First Half of the Third Observing Run
- InceptionTime: Finding AlexNet for Time Series Classification
- GW190425: Observation of a Compact Binary Coalescence with Total Mass
- Effective one-body approach to general relativistic two-body dynamics
- Bilby: A user-friendly Bayesian inference library for gravitational-wave astronomy
- Frequency-domain gravitational waves from non-precessing black-hole binaries. II. A phenomenological model for the advanced detector era
- Robust parameter estimation for compact binaries with ground-based gravitational-wave observations using the LALInference software library
- Observation of gravitational waves from two neutron star-black hole coalescences
- The PyCBC search for gravitational waves from compact binary coalescence
- FINDCHIRP: an algorithm for detection of gravitational waves from inspiraling compact binaries
- An improved effective-one-body model of spinning, nonprecessing binary black holes for the era of gravitational-wave astrophysics with advanced detectors
- An Experimental Review on Deep Learning Architectures for Time Series Forecasting
- Matching post-Newtonian and numerical relativity waveforms: systematic errors and a new phenomenological model for non-precessing black hole binaries
- Towards models of gravitational waveforms from generic binaries II: Modelling precession effects with a single effective precession parameter
- Improved filters for gravitational waves from inspiralling compact binaries
- Time-domain effective-one-body gravitational waveforms for coalescing compact binaries with nonprecessing spins, tides and self-spin effects
- Surrogate model of hybridized numerical relativity binary black hole waveforms
- Phenomenological model for the gravitational-wave signal from precessing binary black holes with two-spin effects
- Bayesian parameter estimation using conditional variational autoencoders for gravitational-wave astronomy
- Including higher order multipoles in gravitational-wave models for precessing binary black holes
- A waveform model for eccentric binary black hole based on effective-one-body-numerical-relativity (EOBNR) formalism
- Comparing Effective-One-Body gravitational waveforms to accurate numerical data
- Gravitational-wave parameter estimation with autoregressive neural network flows
- Nested Sampling with Normalising Flows for Gravitational-Wave Inference
- Frequency domain reduced order model of aligned-spin effective-one-body waveforms with higher-order modes
- Neural Importance Sampling for Rapid and Reliable Gravitational-Wave Inference
- Cosmology and dark energy from joint gravitational wave-GRB observations
- Enabling real-time multi-messenger astrophysics discoveries with deep learning
- MLGWSC-1: The first Machine Learning Gravitational-Wave Search Mock Data Challenge
- Machine Learning Gravitational Waves from Binary Black Hole Mergers
- Denoising of gravitational wave signals via dictionary learning algorithms
- Application of dictionary learning to denoise LIGO's blip noise transients
- Factorization and resummation: A new paradigm to improve gravitational wave amplitudes. III: the spinning test-body terms
- Quasinormal modes of dS and AdS Black Holes: Feedforward neural network method
- Combining effective-one-body accuracy and reduced-order-quadrature speed for binary neutron star merger parameter estimation with machine learning
- Black Holes as Brains: Neural Networks with Area Law Entropy
- Exploring gravitational-wave detection and parameter inference using Deep Learning methods
- Black-Hole evaporation from the perspective of neural networks
- Simulating Transient Noise Bursts in LIGO with gengli
- Peregrine: Sequential simulation-based inference for gravitational wave signals