Deep learning for intermittent gravitational wave signals

TL;DR

This paper explores deep learning methods, including various neural network architectures, to detect and analyze non-Gaussian stochastic gravitational wave backgrounds from unresolved compact binary coalescences, demonstrating promising sensitivity and parameter estimation capabilities.

Contribution

It introduces the first application of deep learning to detect non-Gaussian stochastic GW backgrounds and estimate the astrophysical duty cycle from simulated signals.

Findings

Residual networks achieve sensitivity comparable to traditional statistics.

Deep learning can estimate duty cycle and SNR directly from data.

Classifying duty cycle improves understanding of GW background characteristics.

Abstract

The ensemble of unresolved compact binary coalescences is a promising source of the stochastic gravitational wave (GW) background. For stellar-mass black hole binaries, the astrophysical stochastic GW background is expected to exhibit non-Gaussianity due to their intermittent features. We investigate the application of deep learning to detect such non-Gaussian stochastic GW background and demonstrate it with the toy model employed in Drasco \& Flanagan (2003), in which each burst is described by a single peak concentrated at a time bin. For the detection problem, we compare three neural networks with different structures: a shallower convolutional neural network (CNN), a deeper CNN, and a residual network. We show that the residual network can achieve comparable sensitivity as the conventional non-Gaussian statistic for signals with the astrophysical duty cycle of $\log_{10}\xi \in [-3,-1]$. Furthermore, we apply deep learning for parameter estimation with two approaches, in which the neural network (1) directly provides the duty cycle and the signal-to-noise ratio (SNR) and (2) classifies the data into four classes depending on the duty cycle value. This is the first step of a deep learning application for detecting a non-Gaussian stochastic GW background and extracting information on the astrophysical duty cycle.