First-Order Manifold Data Augmentation for Regression Learning

Ilya Kaufman, Omri Azencot
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:23324-23342, 2024.

Abstract

Data augmentation (DA) methods tailored to specific domains generate synthetic samples by applying transformations that are appropriate for the characteristics of the underlying data domain, such as rotations on images and time warping on time series data. In contrast, domain-independent approaches, e.g. mixup, are applicable to various data modalities, and as such they are general and versatile. While regularizing classification tasks via DA is a well-explored research topic, the effect of DA on regression problems received less attention. To bridge this gap, we study the problem of domain-independent augmentation for regression, and we introduce FOMA: a new data-driven domain-independent data augmentation method. Essentially, our approach samples new examples from the tangent planes of the train distribution. Augmenting data in this way aligns with the network tendency towards capturing the dominant features of its input signals. We evaluate FOMA on in-distribution generalization and out-of-distribution robustness benchmarks, and we show that it improves the generalization of several neural architectures. We also find that strong baselines based on mixup are less effective in comparison to our approach. Our code is publicly available at https://github.com/azencot-group/FOMA

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-kaufman24a, title = {First-Order Manifold Data Augmentation for Regression Learning}, author = {Kaufman, Ilya and Azencot, Omri}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {23324--23342}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/kaufman24a/kaufman24a.pdf}, url = {https://proceedings.mlr.press/v235/kaufman24a.html}, abstract = {Data augmentation (DA) methods tailored to specific domains generate synthetic samples by applying transformations that are appropriate for the characteristics of the underlying data domain, such as rotations on images and time warping on time series data. In contrast, domain-independent approaches, e.g. mixup, are applicable to various data modalities, and as such they are general and versatile. While regularizing classification tasks via DA is a well-explored research topic, the effect of DA on regression problems received less attention. To bridge this gap, we study the problem of domain-independent augmentation for regression, and we introduce FOMA: a new data-driven domain-independent data augmentation method. Essentially, our approach samples new examples from the tangent planes of the train distribution. Augmenting data in this way aligns with the network tendency towards capturing the dominant features of its input signals. We evaluate FOMA on in-distribution generalization and out-of-distribution robustness benchmarks, and we show that it improves the generalization of several neural architectures. We also find that strong baselines based on mixup are less effective in comparison to our approach. Our code is publicly available at https://github.com/azencot-group/FOMA} }
Endnote
%0 Conference Paper %T First-Order Manifold Data Augmentation for Regression Learning %A Ilya Kaufman %A Omri Azencot %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-kaufman24a %I PMLR %P 23324--23342 %U https://proceedings.mlr.press/v235/kaufman24a.html %V 235 %X Data augmentation (DA) methods tailored to specific domains generate synthetic samples by applying transformations that are appropriate for the characteristics of the underlying data domain, such as rotations on images and time warping on time series data. In contrast, domain-independent approaches, e.g. mixup, are applicable to various data modalities, and as such they are general and versatile. While regularizing classification tasks via DA is a well-explored research topic, the effect of DA on regression problems received less attention. To bridge this gap, we study the problem of domain-independent augmentation for regression, and we introduce FOMA: a new data-driven domain-independent data augmentation method. Essentially, our approach samples new examples from the tangent planes of the train distribution. Augmenting data in this way aligns with the network tendency towards capturing the dominant features of its input signals. We evaluate FOMA on in-distribution generalization and out-of-distribution robustness benchmarks, and we show that it improves the generalization of several neural architectures. We also find that strong baselines based on mixup are less effective in comparison to our approach. Our code is publicly available at https://github.com/azencot-group/FOMA
APA
Kaufman, I. & Azencot, O.. (2024). First-Order Manifold Data Augmentation for Regression Learning. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:23324-23342 Available from https://proceedings.mlr.press/v235/kaufman24a.html.

Related Material