Toward learning human-aligned cross-domain robust models by countering misaligned features

Haohan Wang, Zeyi Huang, Hanlin Zhang, Yong Jae Lee, Eric P. Xing
Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence, PMLR 180:2075-2084, 2022.

Abstract

Machine learning has demonstrated remarkable prediction accuracy over i.i.d data, but the accuracy often drops when tested with data from another distribution. In this paper, we aim to offer another view of this problem in a perspective assuming the reason behind this accuracy drop is the reliance of models on the features that are not aligned well with how a data annotator considers similar across these two datasets. We refer to these features as misaligned features. We extend the conventional generalization error bound to a new one for this setup with the knowledge of how the misaligned features are associated with the label. Our analysis offers a set of techniques for this problem, and these techniques are naturally linked to many previous methods in robust machine learning literature. We also compared the empirical strength of these methods demonstrated the performance when these previous techniques are combined, with implementation available.

Cite this Paper

BibTeX
@InProceedings{pmlr-v180-wang22d, title = {Toward learning human-aligned cross-domain robust models by countering misaligned features}, author = {Wang, Haohan and Huang, Zeyi and Zhang, Hanlin and Lee, Yong Jae and Xing, Eric P.}, booktitle = {Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence}, pages = {2075--2084}, year = {2022}, editor = {Cussens, James and Zhang, Kun}, volume = {180}, series = {Proceedings of Machine Learning Research}, month = {01--05 Aug}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v180/wang22d/wang22d.pdf}, url = {https://proceedings.mlr.press/v180/wang22d.html}, abstract = {Machine learning has demonstrated remarkable prediction accuracy over i.i.d data, but the accuracy often drops when tested with data from another distribution. In this paper, we aim to offer another view of this problem in a perspective assuming the reason behind this accuracy drop is the reliance of models on the features that are not aligned well with how a data annotator considers similar across these two datasets. We refer to these features as misaligned features. We extend the conventional generalization error bound to a new one for this setup with the knowledge of how the misaligned features are associated with the label. Our analysis offers a set of techniques for this problem, and these techniques are naturally linked to many previous methods in robust machine learning literature. We also compared the empirical strength of these methods demonstrated the performance when these previous techniques are combined, with implementation available.} }
Endnote
%0 Conference Paper %T Toward learning human-aligned cross-domain robust models by countering misaligned features %A Haohan Wang %A Zeyi Huang %A Hanlin Zhang %A Yong Jae Lee %A Eric P. Xing %B Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2022 %E James Cussens %E Kun Zhang %F pmlr-v180-wang22d %I PMLR %P 2075--2084 %U https://proceedings.mlr.press/v180/wang22d.html %V 180 %X Machine learning has demonstrated remarkable prediction accuracy over i.i.d data, but the accuracy often drops when tested with data from another distribution. In this paper, we aim to offer another view of this problem in a perspective assuming the reason behind this accuracy drop is the reliance of models on the features that are not aligned well with how a data annotator considers similar across these two datasets. We refer to these features as misaligned features. We extend the conventional generalization error bound to a new one for this setup with the knowledge of how the misaligned features are associated with the label. Our analysis offers a set of techniques for this problem, and these techniques are naturally linked to many previous methods in robust machine learning literature. We also compared the empirical strength of these methods demonstrated the performance when these previous techniques are combined, with implementation available.
APA
Wang, H., Huang, Z., Zhang, H., Lee, Y.J. & Xing, E.P.. (2022). Toward learning human-aligned cross-domain robust models by countering misaligned features. Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 180:2075-2084 Available from https://proceedings.mlr.press/v180/wang22d.html.

Related Material