Towards a robust and reliable deep learning approach for detection of compact binary mergers in gravitational wave data

Abstract

The ability of deep learning (DL) approaches to learn generalised signal and noise models, coupled with their fast inference on GPUs, holds great promise for enhancing gravitational-wave (GW) searches in terms of speed, parameter space coverage, and search sensitivity. However, the opaque nature of DL models severely harms their reliability. In this work, we meticulously develop a DL model stage-wise and work towards improving its robustness and reliability. First, we address the problems in maintaining the purity of training data by deriving a new metric that better reflects the visual strength of the 'chirp' signal features in the data. Using a reduced, smooth representation obtained through a variational auto-encoder (VAE), we build a classifier to search for compact binary coalescence (CBC) signals. Our tests on real LIGO data show an impressive performance of the model. However, upon probing the robustness of the model through adversarial attacks, its simple failure modes were identified, underlining how such models can still be highly fragile. As a first step towards bringing robustness, we retrain the model in a novel framework involving a generative adversarial network (GAN). Over the course of training, the model learns to eliminate the primary modes of failure identified by the adversaries. Although absolute robustness is practically impossible to achieve, we demonstrate some fundamental improvements earned through such training, like sparseness and reduced degeneracy in the extracted features at different layers inside the model. We show that these gains are achieved at practically zero loss in terms of model performance on real LIGO data before and after GAN training. Through a direct search on 8.8 days of LIGO data, we recover two significant CBC events from GWTC-2.1, GW190519_153544 and GW190521_074359. We also report the search sensitivity obtained from an injection study.

22 pages, 22 figures

References in corpus (20)

• GW170817: Observation of Gravitational Waves from a Binary Neutron Star Inspiral
• Generative Adversarial Networks
• Advanced Virgo: a 2nd generation interferometric gravitational wave detector
• Advanced LIGO
• GW170814: A Three-Detector Observation of Gravitational Waves from a Binary Black Hole Coalescence
• GW190814: Gravitational Waves from the Coalescence of a 23 M${}_{\odot }$ Black Hole with a 2.6 M${}_{\odot }$ Compact Object
• GW190521: A Binary Black Hole Merger with a Total Mass of $150M_{\odot }$
• GW170608: Observation of a 19-solar-mass Binary Black Hole Coalescence
• Observation of gravitational waves from two neutron star-black hole coalescences
• Gravity Spy: Integrating Advanced LIGO Detector Characterization, Machine Learning, and Citizen Science
• The LIGO Open Science Center
• A Highly Spinning and Aligned Binary Black Hole Merger in the Advanced LIGO First Observing Run
• Noise Reduction in Gravitational-wave Data via Deep Learning
• Classification methods for noise transients in advanced gravitational-wave detectors II: performance tests on Advanced LIGO data
• Detection of gravitational-wave signals from binary neutron star mergers using machine learning
• Searches for Compact Binary Coalescence Events using Neural Networks in LIGO/Virgo Second Observation Period
• Search for binary black hole mergers in the third observing run of Advanced LIGO-Virgo using coherent WaveBurst enhanced with machine learning
• Unmodelled Clustering Methods for Gravitational Wave Populations of Compact Binary Mergers
• A classifier for gravitational-wave inspiral signals in non-ideal single-detector data
• Gravitational-wave signal-to-noise interpolation via neural networks

Cited by in corpus (6)

• Gravity Spy: Lessons Learned and a Path Forward
• New Gravitational Wave Discoveries Enabled by Machine Learning
• Applications of machine learning in gravitational wave research with current interferometric detectors
• A novel stacked hybrid autoencoder for imputing LISA data gaps
• Mass and tidal parameter extraction from gravitational waves of binary neutron stars mergers using deep learning
• Novel Deep Learning Approach to Detecting Binary Black Hole Mergers