Papers

Arbitrary Marginal Neural Ratio Estimation for Simulation-based Inference

F Rozet, G Louppe - arXiv preprint arXiv:2110.00449, 2021 - arxiv.org
Computer Science paper cs.LG Suggest

… This work introduces AMNRE, a novel simulation-based inference method that enables integrationless marginal posterior estimation over arbitrary parameter subspaces. Through our …

Cited by Link to paper

BibTeX

@article{2110.00449v3,
Author        = {François Rozet and Gilles Louppe},
Title         = {Arbitrary Marginal Neural Ratio Estimation for Simulation-based
  Inference},
Eprint        = {2110.00449v3},
ArchivePrefix = {arXiv},
PrimaryClass  = {cs.LG},
Abstract      = {In many areas of science, complex phenomena are modeled by stochastic
parametric simulators, often featuring high-dimensional parameter spaces and
intractable likelihoods. In this context, performing Bayesian inference can be
challenging. In this work, we present a novel method that enables amortized
inference over arbitrary subsets of the parameters, without resorting to
numerical integration, which makes interpretation of the posterior more
convenient. Our method is efficient and can be implemented with arbitrary
neural network architectures. We demonstrate the applicability of the method on
parameter inference of binary black hole systems from gravitational waves
observations.},
Year          = {2021},
Month         = {Oct},
Url           = {http://arxiv.org/abs/2110.00449v3},
File          = {2110.00449v3.pdf}
}

Share