We study the robustness of conformal prediction, a powerful tool for uncertainty quantification, to label noise. Our analysis tackles both regression and classification problems, characterizing when and how it is possible to construct uncertainty sets that correctly cover the unobserved noiseless ground truth labels. Through stylized theoretical examples and practical experiments, we argue that naive conformal prediction covers the noiseless ground truth label unless the noise distribution is adversarially designed. This leads us to believe that correcting for label noise is unnecessary except for pathological data distributions or noise sources. In such cases, we can also correct for noise of bounded size in the conformal prediction algorithm in order to ensure correct coverage of the ground truth labels without score or data regularity.

本研究探究了针对标签噪声的确定性估计方法——“conformal prediction”的鲁棒性。通过理论证明和实验验证，提出了对于正确覆盖未观测噪声的基准真实标签，构建正确的不确定性集合的可能性和条件，并发现除了在数据分布或噪声源方面存在病态的特殊情况外，纠正标签噪声并不必要。在这种情况下，可以在确定性预测算法中纠正有界大小的噪声以确保正确覆盖基准真实标签，而无需直接调整得分或数据稳定性。

预测一致性对标签噪声具有鲁棒性