Nested Sampling with Normalising Flows for Gravitational-Wave Inference

TL;DR

This paper introduces a new nested sampling method using normalising flows within nessai, significantly reducing likelihood evaluations for gravitational-wave parameter estimation while maintaining unbiased results.

Contribution

The novel integration of normalising flows into nested sampling for gravitational-wave inference improves efficiency and reduces likelihood evaluations compared to existing methods.

Findings

Achieved unbiased parameter estimates on simulated gravitational wave signals.

Reduced likelihood evaluations by a factor of 2.07 compared to dynesty.

Demonstrated parallelisable likelihood evaluation in nessai.

Abstract

We present a novel method for sampling iso-likelihood contours in nested sampling using a type of machine learning algorithm known as normalising flows and incorporate it into our sampler nessai. Nessai is designed for problems where computing the likelihood is computationally expensive and therefore the cost of training a normalising flow is offset by the overall reduction in the number of likelihood evaluations. We validate our sampler on 128 simulated gravitational wave signals from compact binary coalescence and show that it produces unbiased estimates of the system parameters. Subsequently, we compare our results to those obtained with dynesty and find good agreement between the computed log-evidences whilst requiring 2.07 times fewer likelihood evaluations. We also highlight how the likelihood evaluation can be parallelised in nessai without any modifications to the algorithm. Finally, we outline diagnostics included in nessai and how these can be used to tune the sampler's settings.