Accelerated Bayesian parameter estimation and model selection for gravitational waves with normalizing flows
This work addresses the computational bottleneck in gravitational wave astrophysics for researchers analyzing gravitational wave data, representing an incremental improvement through integration of existing techniques.
The authors tackled the computational challenge of Bayesian parameter estimation and model selection for gravitational waves by developing an accelerated pipeline using normalizing flows and high-performance computing. Their method achieved consistent results with traditional techniques while reducing computation time by factors of 5× and 15× for 4-dimensional and 11-dimensional problems, respectively.
We present an accelerated pipeline, based on high-performance computing techniques and normalizing flows, for joint Bayesian parameter estimation and model selection and demonstrate its efficiency in gravitational wave astrophysics. We integrate the Jim inference toolkit, a normalizing flow-enhanced Markov chain Monte Carlo (MCMC) sampler, with the learned harmonic mean estimator. Our Bayesian evidence estimates run on $1$ GPU are consistent with traditional nested sampling techniques run on $16$ CPU cores, while reducing the computation time by factors of $5\times$ and $15\times$ for $4$-dimensional and $11$-dimensional gravitational wave inference problems, respectively. Our code is available in well-tested and thoroughly documented open-source packages, ensuring accessibility and reproducibility for the wider research community.