Institute Homepage
Institute Homepage DE Sign In

Back

Empirical Inference Conference Paper 2011

On Causal Discovery with Cyclic Additive Noise Models

Thumb ticker sm dominik 270 cut s
Empirical Inference
Dominik Janzing
  • Research Scientist
Thumb ticker sm l1170153
Empirical Inference
Bernhard Schölkopf
  • Director
no image
Empirical Inference
Joris Mooij

We study a particular class of cyclic causal models, where each variable is a (possibly nonlinear) function of its parents and additive noise. We prove that the causal graph of such models is generically identifiable in the bivariate, Gaussian-noise case. We also propose a method to learn such models from observational data. In the acyclic case, the method reduces to ordinary regression, but in the more challenging cyclic case, an additional term arises in the loss function, which makes it a special case of nonlinear independent component analysis. We illustrate the proposed method on synthetic data.

Author(s): Mooij, J. and Janzing, D. and Schölkopf, B. and Heskes, T.
Book Title: Advances in Neural Information Processing Systems 24
Pages: 639-647
Year: 2011
Day: 0
Editors: J Shawe-Taylor and RS Zemel and PL Bartlett and FCN Pereira and KQ Weinberger
Publisher: Curran Associates, Inc.
Bibtex Type: Conference Paper (inproceedings)
Address: Red Hook, NY, USA
Event Name: Twenty-Fifth Annual Conference on Neural Information Processing Systems (NIPS 2011)
Event Place: Granada, Spain
Digital: 0
Electronic Archiving: grant_archive
Links:

BibTex 

@inproceedings{MooijJSH2012,
  title = {On Causal Discovery with Cyclic Additive Noise Models},
  booktitle = {Advances in Neural Information Processing Systems 24},
  abstract = {We study a particular class of cyclic causal models, where each variable is a (possibly nonlinear) function of its parents and additive noise. We prove that the causal
  graph of such models is generically identifiable in the bivariate, Gaussian-noise case. We also propose a method to learn such models from observational data. In the acyclic case, the method reduces to ordinary regression, but in the more challenging cyclic case, an additional term arises in the loss function, which makes it a special case of nonlinear independent component analysis. We illustrate the proposed method on synthetic data.},
  pages = {639-647},
  editors = {J Shawe-Taylor and RS Zemel and PL Bartlett and FCN Pereira and KQ Weinberger},
  publisher = {Curran Associates, Inc.},
  address = {Red Hook, NY, USA},
  year = {2011},
  slug = {mooijjsh2012},
  author = {Mooij, J. and Janzing, D. and Sch{\"o}lkopf, B. and Heskes, T.}
}