Detecting non-causal artifacts in multivariate linear regression models

Dominik Janzing, Bernhard Schölkopf
Proceedings of the 35th International Conference on Machine Learning, PMLR 80:2245-2253, 2018.

Abstract

We consider linear models where d potential causes X_1,...,X_d are correlated with one target quantity Y and propose a method to infer whether the association is causal or whether it is an artifact caused by overfitting or hidden common causes. We employ the idea that in the former case the vector of regression coefficients has ‘generic’ orientation relative to the covariance matrix Sigma_{XX} of X. Using an ICA based model for confounding, we show that both confounding and overfitting yield regression vectors that concentrate mainly in the space of low eigenvalues of Sigma_{XX}.

Cite this Paper


BibTeX
@InProceedings{pmlr-v80-janzing18a, title = {Detecting non-causal artifacts in multivariate linear regression models}, author = {Janzing, Dominik and Sch{\"o}lkopf, Bernhard}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, pages = {2245--2253}, year = {2018}, editor = {Dy, Jennifer and Krause, Andreas}, volume = {80}, series = {Proceedings of Machine Learning Research}, month = {10--15 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v80/janzing18a/janzing18a.pdf}, url = {https://proceedings.mlr.press/v80/janzing18a.html}, abstract = {We consider linear models where d potential causes X_1,...,X_d are correlated with one target quantity Y and propose a method to infer whether the association is causal or whether it is an artifact caused by overfitting or hidden common causes. We employ the idea that in the former case the vector of regression coefficients has ‘generic’ orientation relative to the covariance matrix Sigma_{XX} of X. Using an ICA based model for confounding, we show that both confounding and overfitting yield regression vectors that concentrate mainly in the space of low eigenvalues of Sigma_{XX}.} }
Endnote
%0 Conference Paper %T Detecting non-causal artifacts in multivariate linear regression models %A Dominik Janzing %A Bernhard Schölkopf %B Proceedings of the 35th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2018 %E Jennifer Dy %E Andreas Krause %F pmlr-v80-janzing18a %I PMLR %P 2245--2253 %U https://proceedings.mlr.press/v80/janzing18a.html %V 80 %X We consider linear models where d potential causes X_1,...,X_d are correlated with one target quantity Y and propose a method to infer whether the association is causal or whether it is an artifact caused by overfitting or hidden common causes. We employ the idea that in the former case the vector of regression coefficients has ‘generic’ orientation relative to the covariance matrix Sigma_{XX} of X. Using an ICA based model for confounding, we show that both confounding and overfitting yield regression vectors that concentrate mainly in the space of low eigenvalues of Sigma_{XX}.
APA
Janzing, D. & Schölkopf, B.. (2018). Detecting non-causal artifacts in multivariate linear regression models. Proceedings of the 35th International Conference on Machine Learning, in Proceedings of Machine Learning Research 80:2245-2253 Available from https://proceedings.mlr.press/v80/janzing18a.html.

Related Material