Robust Optimization for Fairness with Noisy Protected Groups
Robust Optimization for Fairness with Noisy Protected Groups
复制标题
噪声保护组公平性的鲁棒优化
DOI：
--
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
Michael I. Jordan
中科院分区：
文献类型：
--
作者：
S. Wang;Wenshuo Guo;H. Narasimhan;Andrew Cotter;Maya R. Gupta;Michael I. Jordan
许多现有的机器学习公平性标准涉及使一些度量在 extit{protected groups}，例如种族或性别。然而，试图审计或执行这种基于群体的标准的从业人员很容易面临嘈杂或有偏见的受保护群体信息的问题。首先，我们研究了天真地依赖于噪声保护组标签的后果：我们提供了一个上界的公平性违反真正的组$G$时，公平性标准是满足噪声组$hat{G}$。其次，我们介绍了两种新的方法，使用强大的优化，不像天真的方法，只依赖于$hat{G}$，保证满足公平性标准的真正保护组$G$，同时最小化的训练目标。我们提供了理论保证，这样的方法收敛到一个最佳的可行解。使用两个案例研究，我们经验表明，强大的方法实现更好的真正的组公平保证比天真的方法。
Many existing fairness criteria for machine learning involve equalizing some metric across extit{protected groups} such as race or gender. However, practitioners trying to audit or enforce such group-based criteria can easily face the problem of noisy or biased protected group information. First, we study the consequences of na{i}vely relying on noisy protected group labels: we provide an upper bound on the fairness violations on the true groups $G$ when the fairness criteria are satisfied on noisy groups $hat{G}$. Second, we introduce two new approaches using robust optimization that, unlike the na{i}ve approach of only relying on $hat{G}$, are guaranteed to satisfy fairness criteria on the true protected groups $G$ while minimizing a training objective. We provide theoretical guarantees that one such approach converges to an optimal feasible solution. Using two case studies, we show empirically that the robust approaches achieve better true group fairness guarantees than the na{i}ve approach.