Adaptive Conformal Inference Under Distribution Shift

TL;DR

This paper introduces an adaptive conformal inference method that maintains reliable prediction set coverage over time despite unknown and changing data distributions, by continuously re-estimating a key distribution shift parameter.

Contribution

It proposes a novel adaptive framework for conformal inference that works under distribution shift, extending previous methods that assumed data exchangeability.

Findings

Robust prediction sets under distribution shift in real datasets

Achieves desired coverage over long periods despite changing data

Outperforms non-adaptive methods in shifted environments

Abstract

We develop methods for forming prediction sets in an online setting where the data generating distribution is allowed to vary over time in an unknown fashion. Our framework builds on ideas from conformal inference to provide a general wrapper that can be combined with any black box method that produces point predictions of the unseen label or estimated quantiles of its distribution. While previous conformal inference methods rely on the assumption that the data points are exchangeable, our adaptive approach provably achieves the desired coverage frequency over long-time intervals irrespective of the true data generating process. We accomplish this by modelling the distribution shift as a learning problem in a single parameter whose optimal value is varying over time and must be continuously re-estimated. We test our method, adaptive conformal inference, on two real world datasets and find that its predictions are robust to visible and significant distribution shifts.