%0 Journal Article %1 arjovsky2019invariant %A Arjovsky, Martin %A Bottou, Léon %A Gulrajani, Ishaan %A Lopez-Paz, David %D 2019 %K bounds foundations invariance readings theory %T Invariant Risk Minimization %U http://arxiv.org/abs/1907.02893 %X We introduce Invariant Risk Minimization (IRM), a learning paradigm to estimate invariant correlations across multiple training distributions. To achieve this goal, IRM learns a data representation such that the optimal classifier, on top of that data representation, matches for all training distributions. Through theory and experiments, we show how the invariances learned by IRM relate to the causal structures governing the data and enable out-of-distribution generalization.

@article{arjovsky2019invariant, abstract = {We introduce Invariant Risk Minimization (IRM), a learning paradigm to estimate invariant correlations across multiple training distributions. To achieve this goal, IRM learns a data representation such that the optimal classifier, on top of that data representation, matches for all training distributions. Through theory and experiments, we show how the invariances learned by IRM relate to the causal structures governing the data and enable out-of-distribution generalization.}, added-at = {2019-07-12T15:13:48.000+0200}, author = {Arjovsky, Martin and Bottou, Léon and Gulrajani, Ishaan and Lopez-Paz, David}, biburl = {https://www.bibsonomy.org/bibtex/21c0ba7899d38d7588a6757affe4fe1b0/kirk86}, description = {[1907.02893] Invariant Risk Minimization}, interhash = {3e4a1f032a1e88403c7a10c768167ccd}, intrahash = {1c0ba7899d38d7588a6757affe4fe1b0}, keywords = {bounds foundations invariance readings theory}, note = {cite arxiv:1907.02893}, timestamp = {2019-12-26T22:14:24.000+0100}, title = {Invariant Risk Minimization}, url = {http://arxiv.org/abs/1907.02893}, year = 2019 }