Fairness and robustness in anti-causal prediction
Abstract: 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.
License: Creative Commons Attribution 4.0 International (CC BY 4.0)
Submission Length: Regular submission (no more than 12 pages of main content)
Changes Since Last Submission: We have: 1- Corrected a number of typos: "my" -> "by", "excmaple"-> "example", "dsitribution" -> "distribution" as requested by reviewer UgDX 2- Included the details about the M-MMD objective (see equation 4) as requested by reviewers XoML and UgDX 3- Highlighted the fact that C-MMD is the standard method implemented in Tensorflow to enforce equalized odds as requested by reviewer UgDX 4- Added figure 3 in table format in the appendix, and highlighted in the main text that the difference in the robustness of M-MMD and WM-MMD is not statistically significantly different as requested by reviewer QUzX 5- Highlighted that robustness methods in the anticausal settings will not lead to models that satisfy sufficiency as requested by reviewer 9GZG
Assigned Action Editor: ~Jeremie_Mary1
Submission Number: 446