Online Localized Conformal Prediction
This work improves uncertainty quantification for online learning and time-series models, where exchangeability fails, by providing more efficient prediction intervals that adapt to covariate heterogeneity.
Online Localized Conformal Prediction (OLCP) addresses the lack of exchangeability in online and time-series settings by combining online adaptation with covariate-dependent localization, achieving valid long-run coverage with narrower prediction sets than existing methods.
Conformal prediction is a framework that provides valid uncertainty quantification for general models with exchangeable data. However, in the online learning and time-series settings, exchangeability is not satisfied. Existing online conformal methods, such as adaptive conformal inference (ACI), can achieve long-run validity, yet they remain inefficient under covariate heterogeneity because they rely on global calibration. We propose \emph{Online Localized Conformal Prediction (OLCP)}, which combines online adaptation with covariate-dependent localization to better reflect heterogeneity. To reduce sensitivity to the localization bandwidth, we further develop \emph{OLCP-Hedge}, which performs bandwidth selection as an online expert aggregation problem using a constrained online convex optimization framework. Importantly, we provide coverage guarantees for both algorithms and demonstrate through simulations and real-data experiments that the proposed methods attain valid long-run coverage with narrower prediction sets than existing baselines.