Selecting Data Augmentation for Simulating Interventions

Maximilian Ilse, Jakub M Tomczak, Patrick Forré
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:4555-4562, 2021.

Abstract

Machine learning models trained with purely observational data and the principle of empirical risk minimization (Vapnik 1992) can fail to generalize to unseen domains. In this paper, we focus on the case where the problem arises through spurious correlation between the observed domains and the actual task labels. We find that many domain generalization methods do not explicitly take this spurious correlation into account. Instead, especially in more application-oriented research areas like medical imaging or robotics, data augmentation techniques that are based on heuristics are used to learn domain invariant features. To bridge the gap between theory and practice, we develop a causal perspective on the problem of domain generalization. We argue that causal concepts can be used to explain the success of data augmentation by describing how they can weaken the spurious correlation between the observed domains and the task labels. We demonstrate that data augmentation can serve as a tool for simulating interventional data. We use these theoretical insights to derive a simple algorithm that is able to select data augmentation techniques that will lead to better domain generalization.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-ilse21a, title = {Selecting Data Augmentation for Simulating Interventions}, author = {Ilse, Maximilian and Tomczak, Jakub M and Forr{\'e}, Patrick}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {4555--4562}, 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/ilse21a/ilse21a.pdf}, url = {https://proceedings.mlr.press/v139/ilse21a.html}, abstract = {Machine learning models trained with purely observational data and the principle of empirical risk minimization (Vapnik 1992) can fail to generalize to unseen domains. In this paper, we focus on the case where the problem arises through spurious correlation between the observed domains and the actual task labels. We find that many domain generalization methods do not explicitly take this spurious correlation into account. Instead, especially in more application-oriented research areas like medical imaging or robotics, data augmentation techniques that are based on heuristics are used to learn domain invariant features. To bridge the gap between theory and practice, we develop a causal perspective on the problem of domain generalization. We argue that causal concepts can be used to explain the success of data augmentation by describing how they can weaken the spurious correlation between the observed domains and the task labels. We demonstrate that data augmentation can serve as a tool for simulating interventional data. We use these theoretical insights to derive a simple algorithm that is able to select data augmentation techniques that will lead to better domain generalization.} }
Endnote
%0 Conference Paper %T Selecting Data Augmentation for Simulating Interventions %A Maximilian Ilse %A Jakub M Tomczak %A Patrick Forré %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-ilse21a %I PMLR %P 4555--4562 %U https://proceedings.mlr.press/v139/ilse21a.html %V 139 %X Machine learning models trained with purely observational data and the principle of empirical risk minimization (Vapnik 1992) can fail to generalize to unseen domains. In this paper, we focus on the case where the problem arises through spurious correlation between the observed domains and the actual task labels. We find that many domain generalization methods do not explicitly take this spurious correlation into account. Instead, especially in more application-oriented research areas like medical imaging or robotics, data augmentation techniques that are based on heuristics are used to learn domain invariant features. To bridge the gap between theory and practice, we develop a causal perspective on the problem of domain generalization. We argue that causal concepts can be used to explain the success of data augmentation by describing how they can weaken the spurious correlation between the observed domains and the task labels. We demonstrate that data augmentation can serve as a tool for simulating interventional data. We use these theoretical insights to derive a simple algorithm that is able to select data augmentation techniques that will lead to better domain generalization.
APA
Ilse, M., Tomczak, J.M. & Forré, P.. (2021). Selecting Data Augmentation for Simulating Interventions. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:4555-4562 Available from https://proceedings.mlr.press/v139/ilse21a.html.

Related Material