Fairness and robustness in anti-causal prediction
This work addresses fairness and robustness issues for machine learning practitioners in domains like healthcare, but it is incremental as it builds on existing causal and fairness frameworks.
The paper tackles the unclear relationship between fairness and robustness in machine learning by analyzing them through a causal lens in anti-causal prediction tasks, showing that separation fairness and risk invariance robustness are connected, and empirically validating this on a pneumonia detection dataset with improvements in practice.
Robustness to distribution shift and fairness have independently emerged as two important desiderata required of modern machine learning models. While these two desiderata seem related, the connection between them is often unclear in practice. Here, we discuss these connections through a causal lens, focusing on anti-causal prediction tasks, where the input to a classifier (e.g., an image) is assumed to be generated as a function of the target label and the protected attribute. By taking this perspective, we draw explicit connections between a common fairness criterion - separation - and a common notion of robustness - risk invariance. These connections provide new motivation for applying the separation criterion in anticausal settings, and inform old discussions regarding fairness-performance tradeoffs. In addition, our findings suggest that robustness-motivated approaches can be used to enforce separation, and that they often work better in practice than methods designed to directly enforce separation. Using a medical dataset, we empirically validate our findings on the task of detecting pneumonia from X-rays, in a setting where differences in prevalence across sex groups motivates a fairness mitigation. Our findings highlight the importance of considering causal structure when choosing and enforcing fairness criteria.