FIFA: Making Fairness More Generalizable in Classifiers Trained on Imbalanced Data
FIFA: Making Fairness More Generalizable in Classifiers Trained on Imbalanced Data
复制标题
DOI：
10.48550/arxiv.2206.02792
复制
发表时间：
2022-06
期刊：
影响因子：
--
通讯作者：
Zhun Deng;Jiayao Zhang;Linjun Zhang;Ting Ye;Yates Coley;Weijie Su;James Y. Zou
中科院分区：
文献类型：
--
作者：
Zhun Deng;Jiayao Zhang;Linjun Zhang;Ting Ye;Yates Coley;Weijie Su;James Y. Zou
数学公平性在机器学习中起着重要的作用，在学习过程中施加公平性约束是一种常见的方法。然而，许多数据集在某些标签类别中是不平衡的（例如，“健康”）和敏感亚组（例如，“老年患者”）。从经验上讲，这种不平衡不仅会导致分类缺乏可推广性，而且会导致公平性的缺乏，特别是在过度参数化的模型中。例如，公平性感知训练可以确保训练数据上的均衡赔率（EO），但EO远不能满足新用户。在本文中，我们提出了一个理论上的原则，但灵活的方法，即不平衡公平意识（FIFA）。具体来说，FIFA鼓励分类和公平泛化，并且可以灵活地与许多现有的基于logits损失的公平学习方法相结合。虽然我们的主要重点是EO，但FIFA可以直接应用于实现机会均等（EqOpt）;在某些条件下，它也可以应用于其他公平概念。我们通过将FIFA与一个流行的公平分类算法相结合来展示FIFA的强大功能，所产生的算法在几个真实世界的数据集上实现了更好的公平性泛化。
Algorithmic fairness plays an important role in machine learning and imposing fairness constraints during learning is a common approach. However, many datasets are imbalanced in certain label classes (e.g."healthy") and sensitive subgroups (e.g."older patients"). Empirically, this imbalance leads to a lack of generalizability not only of classification, but also of fairness properties, especially in over-parameterized models. For example, fairness-aware training may ensure equalized odds (EO) on the training data, but EO is far from being satisfied on new users. In this paper, we propose a theoretically-principled, yet Flexible approach that is Imbalance-Fairness-Aware (FIFA). Specifically, FIFA encourages both classification and fairness generalization and can be flexibly combined with many existing fair learning methods with logits-based losses. While our main focus is on EO, FIFA can be directly applied to achieve equalized opportunity (EqOpt); and under certain conditions, it can also be applied to other fairness notions. We demonstrate the power of FIFA by combining it with a popular fair classification algorithm, and the resulting algorithm achieves significantly better fairness generalization on several real-world datasets.