2025/09/01 by Jericho Cain, Cain, Jericho
Physics and Astronomy · #Pulsars and Gravitational Waves Research #Cosmology and Gravitation Theories #Gamma-ray bursts and supernovae
paper · pdf · doi:10.1088/1361-6382/ae415e
Abstract Gravitational wave detection requires sophisticated signal processing to identify weak astrophysical signals buried in instrumental noise. Traditional matched filtering approaches face computational challenges with diverse signal morphologies and non-stationary noise. This work presents an unsupervised deep learning methodology integrating continuous wavelet transform (CWT) preprocessing with long short-term memory autoencoder architecture for template-free gravitational wave detection. The CWT provides optimal time–frequency decomposition capturing chirp evolution and transient characteristics essential for compact binary coalescence identification. We train and evaluate our model on LIGO H1 data from Observing Run 4 (O4, 2023–2024), comprising 102 confirmed gravitational wave events from the GWTC-4.0 catalog and 1991 noise segments. During development, we discovered that reconstruction errors from multi-run training (O1–O4) clustered by observing run rather than astrophysical parameters, revealing systematic batch effects from GWOSC’s evolving calibration procedures. Following LIGO’s established practice of per-run optimization, we adopted single-run (O4) training, which eliminated these batch effects and improved recall from 52% to 96% while maintaining 97% precision. The final model achieves exceptional performance on O4 test data: 97.0% precision, 96.1% recall, F 1-score 96.6%, and ROC-AUC 0.994 (102 test signals, 399 noise segments). The reconstruction error distribution shows clean unimodal separation between noise (mean 0.48) and signals (mean 0.77), with only 4 missed detections and 3 false alarms. This unsupervised approach demonstrates that anomaly detection can achieve performance competitive with supervised methods while maintaining template-free operation. While the template-free nature of this approach suggests potential for detecting signals outside current template bank coverage, this capability remains to be validated with exotic signal injections. Our identification and resolution of cross-run batch effects provides methodological guidance for future machine learning applications to multi-epoch gravitational wave datasets.