Unsupervised concept discovery mitigates spurious correlations

Arefin R, Zhang Y, Baratin A, Locatello F, Rish I, Liu D, Kawaguchi K. 2024. Unsupervised concept discovery mitigates spurious correlations. Proceedings of the 41st International Conference on Machine Learning. ICML: International Conference on Machine Learning, PMLR, vol. 235, 1672–1688.

Download (ext.)
Conference Paper | Published | English

Scopus indexed
Author
Arefin, Rifat; Zhang, Yan; Baratin, Aristide; Locatello, FrancescoISTA ; Rish, Irina; Liu, Dianbo; Kawaguchi, Kenji
Department
Series Title
PMLR
Abstract
Models prone to spurious correlations in training data often produce brittle predictions and introduce unintended biases. Addressing this challenge typically involves methods relying on prior knowledge and group annotation to remove spurious correlations, which may not be readily available in many applications. In this paper, we establish a novel connection between unsupervised object-centric learning and mitigation of spurious correlations. Instead of directly inferring subgroups with varying correlations with labels, our approach focuses on discovering concepts: discrete ideas that are shared across input samples. Leveraging existing object-centric representation learning, we introduce CoBalT: a concept balancing technique that effectively mitigates spurious correlations without requiring human labeling of subgroups. Evaluation across the benchmark datasets for sub-population shifts demonstrate superior or competitive performance compared state-of-the-art baselines, without the need for group annotation. Code is available at https://github.com/rarefin/CoBalT
Publishing Year
Date Published
2024-07-30
Proceedings Title
Proceedings of the 41st International Conference on Machine Learning
Publisher
ML Research Press
Acknowledgement
We acknowledge the support of the Canada CIFAR AI Chair Program and IVADO. We thank Mila and Compute Canada for providing computational resources.
Volume
235
Page
1672-1688
Conference
ICML: International Conference on Machine Learning
Conference Location
Vienna, Austria
Conference Date
2024-07-21 – 2024-07-27
eISSN
IST-REx-ID

Cite this

Arefin R, Zhang Y, Baratin A, et al. Unsupervised concept discovery mitigates spurious correlations. In: Proceedings of the 41st International Conference on Machine Learning. Vol 235. ML Research Press; 2024:1672-1688.
Arefin, R., Zhang, Y., Baratin, A., Locatello, F., Rish, I., Liu, D., & Kawaguchi, K. (2024). Unsupervised concept discovery mitigates spurious correlations. In Proceedings of the 41st International Conference on Machine Learning (Vol. 235, pp. 1672–1688). Vienna, Austria: ML Research Press.
Arefin, Rifat, Yan Zhang, Aristide Baratin, Francesco Locatello, Irina Rish, Dianbo Liu, and Kenji Kawaguchi. “Unsupervised Concept Discovery Mitigates Spurious Correlations.” In Proceedings of the 41st International Conference on Machine Learning, 235:1672–88. ML Research Press, 2024.
R. Arefin et al., “Unsupervised concept discovery mitigates spurious correlations,” in Proceedings of the 41st International Conference on Machine Learning, Vienna, Austria, 2024, vol. 235, pp. 1672–1688.
Arefin R, Zhang Y, Baratin A, Locatello F, Rish I, Liu D, Kawaguchi K. 2024. Unsupervised concept discovery mitigates spurious correlations. Proceedings of the 41st International Conference on Machine Learning. ICML: International Conference on Machine Learning, PMLR, vol. 235, 1672–1688.
Arefin, Rifat, et al. “Unsupervised Concept Discovery Mitigates Spurious Correlations.” Proceedings of the 41st International Conference on Machine Learning, vol. 235, ML Research Press, 2024, pp. 1672–88.
All files available under the following license(s):
Copyright Statement:
This Item is protected by copyright and/or related rights. [...]

Link(s) to Main File(s)
Access Level
OA Open Access

Export

Marked Publications

Open Data ISTA Research Explorer

Sources

arXiv 2402.13368

Search this title in

Google Scholar