Justifying Information-Geometric Causal Inference
Machine Learning
2014-02-12 v1
Abstract
Information Geometric Causal Inference (IGCI) is a new approach to distinguish between cause and effect for two variables. It is based on an independence assumption between input distribution and causal mechanism that can be phrased in terms of orthogonality in information space. We describe two intuitive reinterpretations of this approach that makes IGCI more accessible to a broader audience. Moreover, we show that the described independence is related to the hypothesis that unsupervised learning and semi-supervised learning only works for predicting the cause from the effect and not vice versa.
Cite
@article{arxiv.1402.2499,
title = {Justifying Information-Geometric Causal Inference},
author = {Dominik Janzing and Bastian Steudel and Naji Shajarisales and Bernhard Schölkopf},
journal= {arXiv preprint arXiv:1402.2499},
year = {2014}
}
Comments
3 Figures