On Differentiable Bayesian Causal Structure Learning

Simon Rittel; Sebastian Tschiatschek

On Differentiable Bayesian Causal Structure Learning

Simon Rittel, Sebastian Tschiatschek

Published: 05 Jul 2024, Last Modified: 05 Jul 2024Causal@UAI2024 PosterEveryoneRevisionsBibTeXCC BY 4.0

Keywords: Causal Discovery, Structure Learning, Causality, DAGs, Bayesian Probability

TL;DR: Critical review of current Bayesian causal structure learning motivating further research on outlined issues

Abstract: This extended abstract reviews differentiable Bayesian causal structure learning (CSL) and discusses why recent works on Bayesian causal discovery published in top-tier conference do not yet meet important desiderata. In particular, we advocate against the current trend of global regularization via prior terms.

Submission Number: 16

Loading