Improving Gradient-guided Nested Sampling for Posterior Inference
This work addresses the problem of efficient and scalable posterior inference for researchers in Bayesian statistics and machine learning, representing an incremental improvement through the integration of existing methods.
The authors tackled the challenge of scaling nested sampling for posterior inference by developing GGNS, a gradient-guided algorithm that integrates multiple advanced techniques, achieving competitive performance on synthetic and real-world problems and demonstrating faster mode discovery and more accurate partition function estimates when combined with generative flow networks.
We present a performant, general-purpose gradient-guided nested sampling algorithm, ${\tt GGNS}$, combining the state of the art in differentiable programming, Hamiltonian slice sampling, clustering, mode separation, dynamic nested sampling, and parallelization. This unique combination allows ${\tt GGNS}$ to scale well with dimensionality and perform competitively on a variety of synthetic and real-world problems. We also show the potential of combining nested sampling with generative flow networks to obtain large amounts of high-quality samples from the posterior distribution. This combination leads to faster mode discovery and more accurate estimates of the partition function.