EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Residual neural likelihood estimation and its application to gravitational-wave astronomy

Mattia Emma, Gregory Ashton

Physical Review D · 22.06.2026 · 113 (12)

DOI: 10.1103/f5df-cxyg ↗

Kurz gefasst

Fachartikel in der Zeitschrift Physical Review D, online erschienen am 22.06.2026. Verfasst von Mattia Emma und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 123 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Science Foundation, Science and Technology Facilities Council, Cardiff University.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Simulation-based inference provides a powerful framework for Bayesian inference when the likelihood is analytically intractable or computationally prohibitive. By leveraging machine-learning techniques and neural density estimators, it enables flexible likelihood or posterior modeling directly from simulations. We introduce residual neural likelihood estimation (RNLE), a modification of neural likelihood estimation (NLE) that learns the likelihood of non-Gaussian noise in gravitational-wave detector data. Exploiting the additive structure of the signal and noise generation processes, RNLE directly models the noise distribution, substantially reducing the number of simulations required for accurate parameter estimation and improving robustness to realistic noise artifacts. The performance of RNLE is demonstrated using a toy model, simulated gravitational-wave signals, and real detector noise from ground based interferometers. In the presence of loud non-Gaussian transients, we find that RNLE can improve robustness of parameter recovery when trained on appropriately constructed datasets. We further assess the stability of the method by quantifying the variability introduced by retraining the conditional density estimator on statistically identical datasets with different optimization seeds, referred to as training noise, and investigate evidence-weighted combinations of independently trained RNLE realizations as a diagnostic of this variability. An implementation of RNLE is publicly available in the package, enabling its deployment within gravitational-wave astronomy and a broad range of scientific applications requiring flexible, simulation-based likelihood estimation.

Förderung

  • National Science Foundation
  • Science and Technology Facilities Council
  • Cardiff University
  • National MPS Society
  • Australian Research Council
  • Centre National de la Recherche Scientifique
  • Instituto Nazionale di Fisica Nucleare
  • Ministry of Education, Culture, Sports, Science and Technology

Zitieren

APA

Emma, M., & Ashton, G. (2026). Residual neural likelihood estimation and its application to gravitational-wave astronomy. Physical Review D, 113(12). https://doi.org/10.1103/f5df-cxyg

Vancouver

Emma M, Ashton G. Residual neural likelihood estimation and its application to gravitational-wave astronomy. Physical Review D. 2026;113(12). doi:10.1103/f5df-cxyg

BibTeX
@article{Emma2026,
  author = {Emma, Mattia and Ashton, Gregory},
  title = {Residual neural likelihood estimation and its application to gravitational-wave astronomy},
  journal = {Physical Review D},
  year = {2026},
  volume = {113},
  number = {12},
  doi = {10.1103/f5df-cxyg}
}
Kostenlos starten →