When Majorities Prevent Learning: Eliminating Bias to Improve Worst-group and Out-of-distribution Generalization

Yu Yang, Gintare Karolina Dziugaite, Baharan MirzasoleimanICLR 2023 · OpenReview

Abstract

Modern neural networks trained on large datasets have achieved state-of-the-art (in-distribution) generalization performance on various tasks. However, their good generalization performance has been shown to be contributed largely to overfitting spurious biases in large datasets. This is evident by the poor generalization performance of such models on minorities and out-of-distribution data. To alleviate this issue, subsampling the majority groups has been shown to be very effective. However, it is not clear how to find the subgroups (e.g. within a class) in large real-world datasets. Besides, naively subsampling the majority groups can entirely deplete some of their smaller sub-populations and drastically harm the in-distribution performance. Here, we show that tracking gradient trajectories of examples in initial epochs allows for finding large subpopulations of data points. We leverage this observation and propose an importance sampling method that is biased towards selecting smaller subpopulations, and eliminates bias in the large subpopulations. Our experiments confirm the effectiveness of our approach in eliminating spurious biases and learning higher-quality models with superior in- and out-of-distribution performance on various datasets.
BibTeX
@inproceedings{yang2023when,
  author = {Yu Yang and Gintare Karolina Dziugaite and Baharan Mirzasoleiman},
  title = {{When Majorities Prevent Learning: Eliminating Bias to Improve Worst-group and Out-of-distribution Generalization}},
  booktitle = {International Conference on Learning Representations},
  year = {2023},
  url = {https://openreview.net/forum?id=W7udwvFMnAd},
}

Loading related content...

Figure