Dataset Explorer for Learning Summary Statistics

This project provides an interactive dataset explorer for the paper Learning Summary Statistics for Bayesian Inference with Autoencoders.

The app is intended as a lightweight companion to the paper, making it easier to inspect the datasets used in the experiments and understand the structure of the simulation outputs before applying learned summary statistics for Bayesian inference.

Article: Learning Summary Statistics for Bayesian Inference with Autoencoders
Code: enca-inca
Demo: HuggingFace space

Demo

Use the interactive app below to explore the datasets used in the paper.

(Albert et al., 2022)

References

2022

SciPost
Learning summary statistics for bayesian inference with autoencoders

Carlo Albert, Simone Ulzega, Firat Ozdemir, and 2 more authors

SciPost Physics Core, 2022

Abs DOI arXiv Bib HTML

For approximate Bayesian computation with intractable likelihoods, proposes using latent representations from autoencoders as learned summary statistics. The decoder receives explicit/implicit noise so the encoder isolates parameter-relevant signal; validated on two stochastic model classes.
@article{albert2022learning, title = {Learning summary statistics for bayesian inference with autoencoders}, author = {Albert, Carlo and Ulzega, Simone and Ozdemir, Firat and Perez-Cruz, Fernando and Mira, Antonietta}, journal = {SciPost Physics Core}, volume = {5}, number = {3}, pages = {043}, year = {2022}, doi = {10.21468/SciPostPhysCore.5.3.043}, }