Learning Bayesian posteriors with neural networks for gravitational-wave inference

TL;DR

This paper presents a deep learning approach to rapidly produce Bayesian posterior distributions for gravitational-wave source parameters, enabling real-time inference and improved analysis of gravitational-wave data.

Contribution

The authors develop a neural network-based method that approximates Bayesian posteriors directly from data, utilizing reduced-order modeling for efficient data representation.

Findings

Achieves near-instantaneous posterior estimation for gravitational-wave signals.

Demonstrates accuracy comparable to traditional Bayesian inference methods.

Provides open-source code and trained models for community use.

Abstract

We seek to achieve the Holy Grail of Bayesian inference for gravitational-wave astronomy: using deep-learning techniques to instantly produce the posterior $p(\theta|D)$ for the source parameters $\theta$, given the detector data $D$. To do so, we train a deep neural network to take as input a signal + noise data set (drawn from the astrophysical source-parameter prior and the sampling distribution of detector noise), and to output a parametrized approximation of the corresponding posterior. We rely on a compact representation of the data based on reduced-order modeling, which we generate efficiently using a separate neural-network waveform interpolant [A. J. K. Chua, C. R. Galley & M. Vallisneri, Phys. Rev. Lett. 122, 211101 (2019)]. Our scheme has broad relevance to gravitational-wave applications such as low-latency parameter estimation and characterizing the science returns of future experiments. Source code and trained networks are available online at https://github.com/vallis/truebayes.