NeuralEstimators

This repository contains the R interface to the Julia package NeuralEstimators (see here). The package facilitates the user-friendly development of neural point estimators, which are neural networks that transform data into parameter point estimates. They are likelihood free, substantially faster than classical methods, and can be designed to be approximate Bayes estimators. The package caters for any model for which simulation is feasible. See the vignette to get started!

Installation tips

To install the package, please:

1. Install Julia (see here) and R (see here).
2. Install the Julia version of NeuralEstimators.
   • To install from terminal, run the command julia -e 'using Pkg; Pkg.add(url="https://github.com/msainsburydale/NeuralEstimators.jl")'.
3. Install the R interface to NeuralEstimators.
   • Install and load devtools in R and then run devtools::install_github("msainsburydale/NeuralEstimators").

Note that if you wish to simulate training data "on-the-fly" using R functions, you will also need to install the Julia package RCall. Note also that one may compile the vignette during installation (which takes roughly 5 minutes) by adding the argument build_vignettes = TRUE in the final command above.

Supporting and citing

This software was developed as part of academic research. If you would like to support it, please star the repository. If you use it in your research or other activities, please also use the following citation.

@article{,
	author = {Sainsbury-Dale, Matthew and Zammit-Mangion, Andrew and Huser, Raphaël},
	title = {Likelihood-Free Parameter Estimation with Neural {B}ayes Estimators},
	journal = {The American Statistician},
	year = {2024},
	volume = {78},
	pages = {1--14},
	doi = {10.1080/00031305.2023.2249522},
	url = {https://doi.org/10.1080/00031305.2023.2249522}
}

Papers using NeuralEstimators

• Likelihood-free parameter estimation with neural Bayes estimators [paper]
  Matthew Sainsbury-Dale, Andrew Zammit-Mangion, Raphaël Huser (2024)

• Neural Bayes estimators for censored inference with peaks-over-threshold models [paper]
  Jordan Richards, Matthew Sainsbury-Dale, Andrew Zammit-Mangion, Raphaël Huser (2024+)

• Neural Bayes estimators for irregular spatial data using graph neural networks [paper]
  Matthew Sainsbury-Dale, Jordan Richards, Andrew Zammit-Mangion, Raphaël Huser (2024+)

• Modern extreme value statistics for Utopian extremes [paper]
  Jordan Richards, Noura Alotaibi, Daniela Cisneros, Yan Gong, Matheus B. Guerrero, Paolo Redondo, Xuanjie Shao (2023)