Adaptive conformal classification with noisy labels

TL;DR

This paper introduces adaptive conformal prediction methods for classification that effectively handle noisy labels, improving the informativeness and coverage of prediction sets without needing detailed data or model information.

Contribution

It presents new calibration algorithms that adapt to label contamination, providing stronger coverage guarantees and flexibility under various contamination assumptions.

Findings

Enhanced prediction set informativeness under label noise

Robust coverage guarantees in contaminated settings

Validated effectiveness on CIFAR-10H dataset

Abstract

This paper develops novel conformal prediction methods for classification tasks that can automatically adapt to random label contamination in the calibration sample, leading to more informative prediction sets with stronger coverage guarantees compared to state-of-the-art approaches. This is made possible by a precise characterization of the effective coverage inflation (or deflation) suffered by standard conformal inferences in the presence of label contamination, which is then made actionable through new calibration algorithms. Our solution is flexible and can leverage different modeling assumptions about the label contamination process, while requiring no knowledge of the underlying data distribution or of the inner workings of the machine-learning classifier. The advantages of the proposed methods are demonstrated through extensive simulations and an application to object classification with the CIFAR-10H image data set.