Environment Inference for Invariant Learning

Elliot Creager, Joern-Henrik Jacobsen, Richard Zemel
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:2189-2200, 2021.

Abstract

Learning models that gracefully handle distribution shifts is central to research on domain generalization, robust optimization, and fairness. A promising formulation is domain-invariant learning, which identifies the key issue of learning which features are domain-specific versus domain-invariant. An important assumption in this area is that the training examples are partitioned into “domains” or “environments”. Our focus is on the more common setting where such partitions are not provided. We propose EIIL, a general framework for domain-invariant learning that incorporates Environment Inference to directly infer partitions that are maximally informative for downstream Invariant Learning. We show that EIIL outperforms invariant learning methods on the CMNIST benchmark without using environment labels, and significantly outperforms ERM on worst-group performance in the Waterbirds dataset. Finally, we establish connections between EIIL and algorithmic fairness, which enables EIIL to improve accuracy and calibration in a fair prediction problem.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-creager21a, title = {Environment Inference for Invariant Learning}, author = {Creager, Elliot and Jacobsen, Joern-Henrik and Zemel, Richard}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {2189--2200}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/creager21a/creager21a.pdf}, url = {https://proceedings.mlr.press/v139/creager21a.html}, abstract = {Learning models that gracefully handle distribution shifts is central to research on domain generalization, robust optimization, and fairness. A promising formulation is domain-invariant learning, which identifies the key issue of learning which features are domain-specific versus domain-invariant. An important assumption in this area is that the training examples are partitioned into “domains” or “environments”. Our focus is on the more common setting where such partitions are not provided. We propose EIIL, a general framework for domain-invariant learning that incorporates Environment Inference to directly infer partitions that are maximally informative for downstream Invariant Learning. We show that EIIL outperforms invariant learning methods on the CMNIST benchmark without using environment labels, and significantly outperforms ERM on worst-group performance in the Waterbirds dataset. Finally, we establish connections between EIIL and algorithmic fairness, which enables EIIL to improve accuracy and calibration in a fair prediction problem.} }
Endnote
%0 Conference Paper %T Environment Inference for Invariant Learning %A Elliot Creager %A Joern-Henrik Jacobsen %A Richard Zemel %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-creager21a %I PMLR %P 2189--2200 %U https://proceedings.mlr.press/v139/creager21a.html %V 139 %X Learning models that gracefully handle distribution shifts is central to research on domain generalization, robust optimization, and fairness. A promising formulation is domain-invariant learning, which identifies the key issue of learning which features are domain-specific versus domain-invariant. An important assumption in this area is that the training examples are partitioned into “domains” or “environments”. Our focus is on the more common setting where such partitions are not provided. We propose EIIL, a general framework for domain-invariant learning that incorporates Environment Inference to directly infer partitions that are maximally informative for downstream Invariant Learning. We show that EIIL outperforms invariant learning methods on the CMNIST benchmark without using environment labels, and significantly outperforms ERM on worst-group performance in the Waterbirds dataset. Finally, we establish connections between EIIL and algorithmic fairness, which enables EIIL to improve accuracy and calibration in a fair prediction problem.
APA
Creager, E., Jacobsen, J. & Zemel, R.. (2021). Environment Inference for Invariant Learning. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:2189-2200 Available from https://proceedings.mlr.press/v139/creager21a.html.

Related Material