On the pitfalls of Gaussian likelihood scoring for causal discovery
Open Access
- 1 January 2023
- journal article
- research article
- Published by Walter de Gruyter GmbH in Journal of Causal Inference
- Vol. 11 (1)
Abstract
We consider likelihood score-based methods for causal discovery in structural causal models. In particular, we focus on Gaussian scoring and analyze the effect of model misspecification in terms of non-Gaussian error distribution. We present a surprising negative result for Gaussian likelihood scoring in combination with nonparametric regression methods.Keywords
All Related Versions
This publication has 5 references indexed in Scilit:
- Identifying patient-specific root causes with the heteroscedastic noise modelJournal of Computational Science, 2023
- Score-based causal learning in additive noise modelsStatistics, 2015
- CAM: Causal additive models, high-dimensional order search and penalized regressionThe Annals of Statistics, 2014
- Detection of Unfaithfulness and Robust Causal InferenceMinds and Machines, 2008
- Causation, Prediction, and Search2001