Previous works (Bojanowski et al., 2018; Berthelot et al., 2018) have shown that interpolations between a pair of latent features correspond to semantically meaningful, smooth interpolation in the input space.
By constructing the path in the latent space, we can better capture the semantic changes while traveling between the sensitive groups and hence result in a better fairness regularizer that we coinfair mixup.
References
Verma, Vikas, et al. "Manifold mixup: Better representations by interpolating hidden states." International Conference on Machine Learning. PMLR, 2019.
用 mixup 增强 group fairness metrics 的泛化性,通过两个 sensitive group 样例的渐近插值,搭建一个逐步的变换过程来连接彼此(图示很明白),帮助模型学习的更公平。以往的方法都依赖于数据,本文方法与数据无关,可跨模态、跨数据等等。搞了点理论证明,在表格、视觉和文本的 fairness 数据集上实验了,有效果。
总的来说,就是在二分类上用了mixup,没啥借鉴意义,可以学学证明。
信息
1 学习到的新东西:
2 通过Related Work了解到了哪些知识
3 实验验证任务,如果不太熟悉,需要简单描述
搞了几个模态的二分类任务,计算平均准确率 (average precision, AP) 和 两个 fairness 的指标 (demographic parity, DP; equalized odds, EO)。
4 在你认知范围内,哪些其它任务可以尝试
无,本文针对特定任务和指标应用了mixup,无改动。
5 好的句子
以句子为单位收集,读起来不错,觉得有机会用上,就摘抄。
References