[Suggestion] New papers

[andrewzm]

Title: Neural Methods for Amortized Inference

Short description: Review paper on methods for doing amortized inference with neural networks, from point estimation, to approximate Bayesian inference, to summary-statistic construction and likelihood approximation.

Link to paper: https://arxiv.org/abs/2404.12484

@Article{zammit2024neural,
title={Neural Methods for Amortised Inference},
author={Zammit-Mangion, Andrew and Sainsbury-Dale, Matthew and Huser, Rapha{"e}l},
journal={arXiv preprint arXiv:2404.12484},
year={2024},
note={in press with Annual Review of Statistics and Its Application}
}

---

Title: Likelihood-Free Parameter Estimation with Neural Bayes Estimators

Short description: Amortized point estimation using neural networks from a decision theoretic perspective. Uses DeepSets for estimating parameters from conditionally independent replicates.

Link to paper: https://www.tandfonline.com/doi/full/10.1080/00031305.2023.2249522#abstract

@Article{sainsbury2024likelihood,
title={Likelihood-free parameter estimation with neural {B}ayes estimators},
author={Sainsbury-Dale, Matthew and Zammit-Mangion, Andrew and Huser, Rapha{"e}l},
journal={The American Statistician},
volume={78},
number={1},
pages={1--14},
year={2024},
publisher={Taylor & Francis}
}

---

Title: Neural Bayes Estimators for Irregular Spatial Data using Graph Neural Networks

Short description: Proposes graph neural networks to make amortized inference of parameters in spatial process models from irregularly spaced data.

Link to paper: https://arxiv.org/abs/2310.02600

@Article{sainsbury2023neural,
title={Neural {B}ayes estimators for irregular spatial data using graph neural networks},
author={Sainsbury-Dale, Matthew and Richards, Jordan and Zammit-Mangion, Andrew and Huser, Rapha{"e}l},
journal={arXiv preprint arXiv:2310.02600},
year={2023}
}

---

Title: Neural Bayes Estimators for Censored Inference with Peaks-Over-Threshold Models

Short description: Proposes neural networks for making inference on parameters of peaks-over-threshold spatial models from censored data.

Link to paper: https://arxiv.org/abs/2306.15642

@Article{richards2023likelihood,
title={Likelihood-free neural {B}ayes estimators for censored peaks-over-threshold models},
author={Richards, Jordan and Sainsbury-Dale, Matthew and Zammit-Mangion, Andrew and Huser, Rapha{"e}l},
journal={arXiv preprint arXiv:2306.15642},
year={2023}
}