Training Strategies for Deep Learning Gravitational-Wave Searches

Abstract

Compact binary systems emit gravitational radiation which is potentially detectable by current Earth bound detectors. Extracting these signals from the instruments' background noise is a complex problem and the computational cost of most current searches depends on the complexity of the source model. Deep learning may be capable of finding signals where current algorithms hit computational limits. Here we restrict our analysis to signals from non-spinning binary black holes and systematically test different strategies by which training data is presented to the networks. To assess the impact of the training strategies, we re-analyze the first published networks and directly compare them to an equivalent matched-filter search. We find that the deep learning algorithms can generalize low signal-to-noise ratio (SNR) signals to high SNR ones but not vice versa. As such, it is not beneficial to provide high SNR signals during training, and fastest convergence is achieved when low SNR samples are provided early on. During testing we found that the networks are sometimes unable to recover any signals when a false alarm probability $<10^{-3}$ is required. We resolve this restriction by applying a modification we call unbounded Softmax replacement (USR) after training. With this alteration we find that the machine learning search retains $\ge 91.5\%$ of the sensitivity of the matched-filter search down to a false-alarm rate of 1 per month.

18 pages, 12 figures, 3 tables, supplemental materials at https://github.com/gwastro/ml-training-strategies

References in corpus (16)

• GW170817: Observation of Gravitational Waves from a Binary Neutron Star Inspiral
• Multi-messenger Observations of a Binary Neutron Star Merger
• Advanced LIGO
• An Ordinary Short Gamma-Ray Burst with Extraordinary Implications: Fermi-GBM Detection of GRB 170817A
• INTEGRAL Detection of the First Prompt Gamma-Ray Signal Coincident with the Gravitational Wave Event GW170817
• An improved effective-one-body model of spinning, nonprecessing binary black holes for the era of gravitational-wave astrophysics with advanced detectors
• The Electromagnetic Counterpart of the Binary Neutron Star Merger LIGO/Virgo GW170817. I. Dark Energy Camera Discovery of the Optical Counterpart
• 3-OGC: Catalog of gravitational waves from compact-binary mergers
• Nested Sampling with Normalising Flows for Gravitational-Wave Inference
• Detection of gravitational-wave signals from binary neutron star mergers using machine learning
• Machine Learning Gravitational Waves from Binary Black Hole Mergers
• Deep Learning Ensemble for Real-time Gravitational Wave Detection of Spinning Binary Black Hole Mergers
• Deep-Learning Continuous Gravitational Waves: Multiple detectors and realistic noise
• The GstLAL Search Analysis Methods for Compact Binary Mergers in Advanced LIGO's Second and Advanced Virgo's First Observing Runs
• Improved deep learning techniques in gravitational-wave data analysis
• Advances in Machine and Deep Learning for Modeling and Real-time Detection of Multi-Messenger Sources

Cited by in corpus (25)

• MLGWSC-1: The first Machine Learning Gravitational-Wave Search Mock Data Challenge
• Deep Residual Networks for Gravitational Wave Detection
• From One to Many: A Deep Learning Coincident Gravitational-Wave Search
• Deep Learning Detection and Classification of Gravitational Waves from Neutron Star-Black Hole Mergers
• Inference-optimized AI and high performance computing for gravitational wave detection at scale
• Ensemble of Deep Convolutional Neural Networks for real-time gravitational wave signal recognition
• Identification of Binary Neutron Star Mergers in Gravitational-Wave Data Using YOLO One-Shot Object Detection
• New Gravitational Wave Discoveries Enabled by Machine Learning
• Applications of machine learning in gravitational wave research with current interferometric detectors
• Convolutional neural network search for long-duration transient gravitational waves from glitching pulsars
• Premerger detection of massive black hole binaries using deep learning
• Neural network time-series classifiers for gravitational-wave searches in single-detector periods
• Matched filtering for gravitational wave detection without template bank driven by deep learning template prediction model bank
• Comparative study of 1D and 2D convolutional neural network models with attribution analysis for gravitational wave detection from compact binary coalescences
• Rapid identification of time-frequency domain gravitational wave signals from binary black holes using deep learning
• Novel neural-network architecture for continuous gravitational waves
• Convolutional Neural Networks for signal detection in real LIGO data
• Binary Neutron Star Merger Search Pipeline Powered by Deep Learning
• Novel Deep Learning Approach to Detecting Binary Black Hole Mergers
• 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
• Large-kernel Convolutional Neural Networks for Wide Parameter-Space Searches of Continuous Gravitational Waves
• Physics-inspired spatiotemporal-graph AI ensemble for the detection of higher order wave mode signals of spinning binary black hole mergers
• Transformer Networks for Continuous Gravitational-wave Searches
• Search for exotic gravitational wave signals beyond general relativity using deep learning
• Constraints on typical relic gravitational waves based on data of LIGO