Gravitational-wave parameter estimation with autoregressive neural network flows

Abstract

We introduce the use of autoregressive normalizing flows for rapid likelihood-free inference of binary black hole system parameters from gravitational-wave data with deep neural networks. A normalizing flow is an invertible mapping on a sample space that can be used to induce a transformation from a simple probability distribution to a more complex one: if the simple distribution can be rapidly sampled and its density evaluated, then so can the complex distribution. Our first application to gravitational waves uses an autoregressive flow, conditioned on detector strain data, to map a multivariate standard normal distribution into the posterior distribution over system parameters. We train the model on artificial strain data consisting of IMRPhenomPv2 waveforms drawn from a five-parameter $(m_1,m_2,{\varphi }_0,t_c,d_L)$ prior and stationary Gaussian noise realizations with a fixed power spectral density. This gives performance comparable to current best deep-learning approaches to gravitational-wave parameter estimation. We then build a more powerful latent variable model by incorporating autoregressive flows within the variational autoencoder framework. This model has performance comparable to Markov chain Monte Carlo and, in particular, successfully models the multimodal ${\varphi }_0$ posterior. Finally, we train the autoregressive latent variable model on an expanded parameter space, including also aligned spins $({\chi }_{1z},{\chi }_{2z})$ and binary inclination ${\theta }_{JN}$, and show that all parameters and degeneracies are well-recovered. In all cases, sampling is extremely fast, requiring less than two seconds to draw $10^4$ posterior samples.

arXiv comment14 pages, 7 figures

References in corpus (23)

• Adam: A Method for Stochastic Optimization
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Auto-Encoding Variational Bayes
• emcee: The MCMC Hammer
• Multi-messenger Observations of a Binary Neutron Star Merger
• WaveNet: A Generative Model for Raw Audio
• 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
• Stochastic Backpropagation and Approximate Inference in Deep Generative Models
• Conditional Image Generation with PixelCNN Decoders
• Variational Inference with Normalizing Flows
• Properties of the binary neutron star merger GW170817
• Weight Uncertainty in Neural Networks
• 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
• Pyro: Deep Universal Probabilistic Programming
• A guide to LIGO-Virgo detector noise and extraction of transient gravitational-wave signals
• MADE: Masked Autoencoder for Distribution Estimation
• Bayesian parameter estimation using conditional variational autoencoders for gravitational-wave astronomy
• Cyclical Annealing Schedule: A Simple Approach to Mitigating KL Vanishing
• Learning Bayesian posteriors with neural networks for gravitational-wave inference
• Decontamination of Mutual Contamination Models

Cited by in corpus (98)

• Real-time gravitational-wave science with neural posterior estimation
• Bayesian parameter estimation using conditional variational autoencoders for gravitational-wave astronomy
• Nested Sampling with Normalising Flows for Gravitational-Wave Inference
• Neural Importance Sampling for Rapid and Reliable Gravitational-Wave Inference
• Parameter estimation with gravitational waves
• The Science of the Einstein Telescope
• Gwbench: a novel Fisher information package for gravitational-wave benchmarking
• Bilby-MCMC: An MCMC sampler for gravitational-wave inference
• Accelerating parameter estimation of gravitational waves from compact binary coalescence using adaptive frequency resolutions
• Real-time gravitational-wave inference for binary neutron stars using machine learning
• Importance nested sampling with normalising flows
• Neural Posterior Estimation with guaranteed exact coverage: the ringdown of GW150914
• Training Strategies for Deep Learning Gravitational-Wave Searches
• Extraction of Binary Black Hole Gravitational Wave Signals from Detector Data Using Deep Learning
• Convolutional neural networks for the detection of the early inspiral of a gravitational-wave signal
• Normalizing flows as an avenue to study overlapping gravitational wave signals
• Dynamical mass inference of galaxy clusters with neural flows
• LtU-ILI: An All-in-One Framework for Implicit Inference in Astrophysics and Cosmology
• VARAHA: A Fast Non-Markovian sampler for estimating Gravitational-Wave posteriors
• SBI++: Flexible, Ultra-fast Likelihood-free Inference Customized for Astronomical Applications
• Generalized Approach to Matched Filtering using Neural Networks
• Inference-optimized AI and high performance computing for gravitational wave detection at scale
• LSTM and CNN application for core-collapse supernova search in gravitational wave real data
• Gravitational wave populations and cosmology with neural posterior estimation
• An Infrared Search for Kilonovae with the WINTER Telescope. I. Binary Neutron Star Mergers
• Improving significance of binary black hole mergers in Advanced LIGO data using deep learning : Confirmation of GW151216
• Physics-inspired deep learning to characterize the signal manifold of quasi-circular, spinning, non-precessing binary black hole mergers
• Deep Generative Models of Gravitational Waveforms via Conditional Autoencoder
• Applications of machine learning in gravitational wave research with current interferometric detectors
• Lightning-Fast Gravitational Wave Parameter Inference through Neural Amortization
• Adapting to noise distribution shifts in flow-based gravitational-wave inference
• Merger-Ringdown Consistency: A New Test of Strong Gravity using Deep Learning
• Unbiased likelihood-free inference of the Hubble constant from light standard sirens
• Density estimation with Gaussian processes for gravitational-wave posteriors
• Generalised gravitational burst generation with Generative Adversarial Networks
• PyROQ: a Python-based Reduced Order Quadrature Building Code for Fast Gravitational Wave Inference
• Rapid parameter estimation for an all-sky continuous gravitational wave search using conditional varitational auto-encoders
• Advancing Space-Based Gravitational Wave Astronomy: Rapid Parameter Estimation via Normalizing Flows
• Interpretable AI forecasting for numerical relativity waveforms of quasi-circular, spinning, non-precessing binary black hole mergers
• Efficient Reduced Order Quadrature Construction Algorithms for Fast Gravitational Wave Inference
• Convolutional neural network for gravitational-wave early alert: Going down in frequency
• Advances in Machine and Deep Learning for Modeling and Real-time Detection of Multi-Messenger Sources
• Strong-lensing source reconstruction with variationally optimised Gaussian processes
• Prompt sky localization of compact binary sources using a meshfree approximation
• Bayesian evidence estimation from posterior samples with normalizing flows
• Nonparametric Representation of Neutron Star Equation of State Using Variational Autoencoder
• Simulation-based inference of black hole ringdowns in the time domain
• Towards a robust and reliable deep learning approach for detection of compact binary mergers in gravitational wave data
• Rapid identification of time-frequency domain gravitational wave signals from binary black holes using deep learning
• Parameter estimation of microlensed gravitational waves with Conditional Variational Autoencoders
• Comparison of Affine and Rational Quadratic Spline Coupling and Autoregressive Flows through Robust Statistical Tests
• Classifying binary black holes from Population III stars with the Einstein Telescope: A machine-learning approach
• Assessing the impact of non-Gaussian noise on convolutional neural networks that search for continuous gravitational waves
• Computational Techniques for Parameter Estimation of Gravitational Wave Signals
• Deep learning for intermittent gravitational wave signals
• Conditional Noise Deep Learning for Parameter Estimation of Gravitational Wave Events
• Pinpointing coalescing binary neutron star sources with the IGWN, including LIGO-Aundha
• An approach to cold dark matter deviation and the $H_0$ tension problem by using machine learning
• Mass and tidal parameter extraction from gravitational waves of binary neutron stars mergers using deep learning
• Deep learning to detect gravitational waves from binary close encounters: Fast parameter estimation using normalizing flows
• A novel stacked hybrid autoencoder for imputing LISA data gaps
• Convolutional Neural Networks for signal detection in real LIGO data
• Simulation-based Inference for Gravitational-waves from Intermediate-Mass Binary Black Holes in Real Noise
• Inferring small neutron star spins with neutron star-black hole mergers
• Deep-Learning Classification and Parameter Inference of Rotational Core-Collapse Supernovae
• Inferring Binary Properties from Gravitational Wave Signals
• Parameter inference for coalescing massive black hole binaries using deep learning
• Waging a Campaign: Results from an Injection-Recovery Study involving 35 numerical Relativity Simulations and three Waveform Models
• AI and extreme scale computing to learn and infer the physics of higher order gravitational wave modes of quasi-circular, spinning, non-precessing binary black hole mergers
• Fast and accurate parameter estimation of high-redshift sources with the Einstein Telescope
• Comparison of neural network architectures for feature extraction from binary black hole merger waveforms
• Optimized localization for gravitational-waves from merging binaries
• Nonlinear noise regression in gravitational-wave detectors with convolutional neural networks
• Robust, Rapid, and Simple Gravitational-wave Parameter Estimation
• Detecting the third family of compact stars with normalizing flows
• Generative adversarial network for stellar core-collapse gravitational waves
• Unifying supervised learning and VAEs -- coverage, systematics and goodness-of-fit in normalizing-flow based neural network models for astro-particle reconstructions
• Accelerated inference of microlensed gravitational waves with machine learning
• Discovering gravitational waveform distortions from lensing: A deep dive into GW231123
• Boosting the Efficiency of Parametric Detection with Hierarchical Neural Networks
• Rapid Parameter Estimation for Merging Massive Black Hole Binaries Using Continuous Normalizing Flows
• Extracting overlapping gravitational-wave signals of galactic compact binaries: a mini review
• Flexible Gravitational-Wave Parameter Estimation with Transformers
• Statistically-informed deep learning for gravitational wave parameter estimation
• The impact of precession and higher-order multipoles for gravitational wave cosmological inference
• Comparing next-generation detector configurations for high-redshift gravitational wave sources with neural posterior estimation
• Genetic-algorithm-optimized neural networks for gravitational wave classification
• Accurate and efficient simulation-based inference for massive black-hole binaries with LISA
• Legacy of the First Workshop on Gravitational Wave Astrophysics for Early Career Scientists
• labrador: A domain-optimized machine-learning tool for gravitational wave inference
• Binary Black Hole Parameter Estimation with Hybrid CNN-Transformer Neural Networks
• Denoising gravitational wave with deep learning in the time-frequency domain
• Template-Free Gravitational Wave Detection with CWT-LSTM Autoencoders: A Case Study of Run-Dependent Calibration Effects in LIGO Data
• Lightweight posterior construction for gravitational-wave catalogs with the Kolmogorov-Arnold network
• Assessment of normalizing flows for parameter estimation on time-frequency representations of gravitational-wave data
• Parameter inference of millilensed gravitational waves using neural spline flows
• Sequence modeling of higher-order wave modes of binary black hole mergers
• Reconsidering the consistent use of precessing, higher order multipole models for gravitational wave analyses