Affiliation:
1. Seminar for Statistics, Department of Mathematics , ETH Zürich , Switzerland
Abstract
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.
Subject
Statistics, Probability and Uncertainty,Statistics and Probability
Reference12 articles.
1. Spirtes P, Glymour CN, Scheines R, Heckerman D. Causation, prediction, and search. Cambridge, MA, USA: MIT Press; 2000.
2. Hoyer P, Janzing D, Mooij JM, Peters J, Schölkopf B. Nonlinear causal discovery with additive noise models. Adv Neur In. 2008;21:689–96.
3. Peters J, Mooij J, Janzing D, Schölkopf B. Causal discovery with continuous additive noise models. J Mach Learn Res. 2014;15:2009–53.
4. Nowzohour C, Bühlmann P. Score-based causal learning in additive noise models. Statistics. 2016;50(3):471–85.
5. Bühlmann P, Peters J, Ernest J. CAM: Causal additive models, high-dimensional order search and penalized regression. Ann Statist. 2014;42(6):2526–56.