Nested sampling with normalizing flows for gravitational-wave inference

Abstract

We present a novel method for sampling iso-likelihood contours in nested sampling using a type of machine learning algorithm known as normalising flows and incorporate it into our sampler Nessai. Nessai is designed for problems where computing the likelihood is computationally expensive and therefore the cost of training a normalising flow is o↵set by the overall reduction in the number of likelihood evaluations. We validate our sampler on 128 simulated gravitational wave signals from compact binary coalescence and show that it produces unbiased estimates of the system parameters. Subsequently, we compare our results to those obtained with dynesty and find good agreement between the computed log-evidences whilst requiring 2.07 times fewer likelihood evaluations. We also highlight how the likelihood evaluation can be parallelised in Nessai without any modifications to the algorithm. Finally, we outline diagnostics included in Nessai and how these can be used to tune the sampler's settings.

The last three years: multiband gravitational-wave observations of stellar-mass binary black holes

“…Parameter estimation -We perform fully Bayesian inference on simulated data, injecting the signal in zero noise data. We use the Balrog codesuite, which provides a rigid adiabatic approximation response model for LISA [71], interfaces to a number of stochastic samplers [72][73][74][75], time-and frequency-domain waveforms for double white dwarfs and stellar mass binary black holes, infrastructure for time-delay-interferometry [76], and access to the LDC codebase under development within the LISA Consortium [61]. Following closely the methodology in [19], we employ a Gaussian likelihood on the time delay interferometry (TDI) variables A, E, T , constructed as noise orthogonal combination of the Michelson TDI variables X, Y, Z.…”

Bilby-MCMC: an MCMC sampler for gravitational-wave inference

“…The normalizing flows class of machine learning algorithms (Papamakarios et al 2019) learn a bijective map from the target density (the set of training samples drawn from the MCMC sampler) to a latent space, in our case a multivariate Gaussian. Normalizing flows have previously been used in gravitational-wave astronomy to directly sample the CBC posterior distribution (Green et al 2020;Green & Gair 2021) and as way to propose new points in a nested sampler (Williams, Veitch & Messenger 2021). Following the work of Hoffman et al (2019), Moss (2020), we use the NFLOWS package (Durkan et al 2020), which implements the normalizing flows algorithm in PYTORCH (Paszke et al 2019), to learn the proposal distribution.…”

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

“…This would be especially beneficial for searches that require low latency, such as the early warning of binary neutron star mergers (Baltus et al, 2021 ; Yu et al, 2021 ). Other successful usage of ML techniques in GW astronomy include the identification of various GW events (Bayley et al, 2020 ; Chan et al, 2020 ; Dreissigacker and Prix, 2020 ; Huerta et al, 2020 ; Krastev, 2020 ; Schäfer et al, 2020 ; Wong et al, 2020 ; Beheshtipour and Papa, 2021 ; Chang et al, 2021 ; Chatterjee et al, 2021 ; López et al, 2021 ; Marianer et al, 2021 ; Mishra et al, 2021 ; Saiz-Pérez et al, 2021 ; Wei and Huerta, 2021 ; Yan et al, 2021 ), source parameter estimations (Gabbard et al, 2019 ; Chatterjee et al, 2020 ; Chua and Vallisneri, 2020 ; Green et al, 2020 ; Talbot and Thrane, 2020 ; Álvares et al, 2021 ; D'Emilio et al, 2021 ; Krastev et al, 2021 ; Williams et al, 2021 ; Xia et al, 2021 ), and detector characterization (Biswas et al, 2020 ; Colgan et al, 2020 ; Cuoco et al, 2020 ; Essick et al, 2020 ; Torres-Forné et al, 2020 ; Mogushi, 2021 ; Sankarapandian and Kulis, 2021 ; Soni et al, 2021 ; Zhan et al, 2021 ). Besides GW astronomy, the usage of CNNs has led to breakthroughs in a variety of topics related to time-series forecasting and classification (e.g., Refs.…”