Subgroup Robustness Grows On Trees: An Empirical Baseline Investigation
Josh Gardner, Zoran Popovi\'c, Ludwig Schmidt

TL;DR
This paper empirically demonstrates that tree-based models exhibit strong subgroup robustness and stability across metrics, outperforming many fairness and robustness methods on tabular data, and serve as effective baselines.
Contribution
It provides a comprehensive empirical comparison showing tree-based models' superior subgroup robustness and stability, establishing them as effective baselines for fair and robust learning on tabular data.
Findings
Tree-based models have strong subgroup robustness.
Tree models are less sensitive to hyperparameters.
Tree models are less costly to train.
Abstract
Researchers have proposed many methods for fair and robust machine learning, but comprehensive empirical evaluation of their subgroup robustness is lacking. In this work, we address this gap in the context of tabular data, where sensitive subgroups are clearly-defined, real-world fairness problems abound, and prior works often do not compare to state-of-the-art tree-based models as baselines. We conduct an empirical comparison of several previously-proposed methods for fair and robust learning alongside state-of-the-art tree-based methods and other baselines. Via experiments with more than model configurations on eight datasets, we show that tree-based methods have strong subgroup robustness, even when compared to robustness- and fairness-enhancing methods. Moreover, the best tree-based models tend to show good performance over a range of metrics, while robust or group-fair…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
Taxonomy
TopicsEthics and Social Impacts of AI · Privacy-Preserving Technologies in Data · Explainable Artificial Intelligence (XAI)
