Training individually fair ML models with sensitive subspace robustness
Training individually fair ML models with sensitive subspace robustness
复制标题
DOI：
--
复制
发表时间：
2019-06
期刊：
影响因子：
--
通讯作者：
M. Yurochkin;Amanda Bower;Yuekai Sun
中科院分区：
文献类型：
--
作者：
M. Yurochkin;Amanda Bower;Yuekai Sun
我们提出了一种训练机器学习模型的方法，这种方法是公平的，因为它们的性能在特征的某些扰动下保持不变。例如，简历筛选系统的性能应该在申请人姓名发生变化的情况下保持不变。我们通过将这种直观的公平概念与 Dwork 等人提出的个人公平的原始概念联系起来，将其形式化，并表明所提出的方法实现了这种公平概念。我们还证明了该方法在两项容易受到性别和种族偏见影响的机器学习任务上的有效性。
We propose an approach to training machine learning models that are fair in the sense that their performance is invariant under certain perturbations to the features. For example, the performance of a resume screening system should be invariant under changes to the name of the applicant. We formalize this intuitive notion of fairness by connecting it to the original notion of individual fairness put forth by Dwork et al and show that the proposed approach achieves this notion of fairness. We also demonstrate the effectiveness of the approach on two machine learning tasks that are susceptible to gender and racial biases.