跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2605.11490 · v3

Adaptive Calibration in Non-Stationary Environments

首次提交 2026-05-12 12:06
版本更新 2026-11-13 12:06
雷达观测 2026-11-17 04:57
arXiv ↗PDF ↗

摘要

Making calibrated online predictions is a central challenge in modern AI systems. Much of the existing literature focuses on fully adversarial environments where outcomes may be arbitrary, leading to conservative algorithms that can perform suboptimally in more benign settings, such as when outcomes are nearly stationary. This gap raises a natural question: can we design online prediction algorithms whose calibration error automatically adapts to the degree of non-stationarity in the environment, smoothly interpolating between i.i.d. and adversarial regimes? We answer this question in the affirmative and develop a suite of algorithms that achieve adaptive calibration guarantees under multiple calibration measures. Specifically, with TT being the number of rounds, WW being the unknown number of i.i.d. segments of the environment, and C∈[0,T]C\in[0,T] being another unknown non-stationary measure defined as the minimal ℓ1\ell_1 deviation of the mean outcomes, our algorithms attain O~(min⁡{T+(TC)13,WT})\widetilde{\mathcal{O}}(\min\{\sqrt{T}+(TC)^{\frac{1}{3}}, \sqrt{WT}\}) for ℓ1\ell_1 calibration error and O~(min⁡{(1+C)13,W})\widetilde{\mathcal{O}}(\min\{(1+C)^{\frac{1}{3}}, W\}) for both ℓ2\ell_2 and pseudo KL calibration error. These bounds match the optimal rates in the stationary case (C=0C=0 and W=1W=1) and recover known guarantees in the fully adversarial regime (C,W=Ω(T)C, W=Ω(T)). Our approach builds on and extends prior work (Hu et al., 2026; Luo et al., 2025), introducing an epoch-based scheduling together with a novel non-uniform partition of the prediction space that allocates finer resolution near the underlying ground truth.

作者备注:50 pages, NeurIPS 2026

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

本站首次收录为 v3,暂无更早版本可对照。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v3 · 2026-11-13 12:06 · 最新观测版本

Adaptive Calibration in Non-Stationary Environments

Making calibrated online predictions is a central challenge in modern AI systems. Much of the existing literature focuses on fully adversarial environments where outcomes may be arbitrary, leading to conservative algorithms that can perform suboptimally in more benign settings, such as when outcomes are nearly stationary. This gap raises a natural question: can we design online prediction algorithms whose calibration error automatically adapts to the degree of non-stationarity in the environment, smoothly interpolating between i.i.d. and adversarial regimes? We answer this question in the affirmative and develop a suite of algorithms that achieve adaptive calibration guarantees under multiple calibration measures. Specifically, with TT being the number of rounds, WW being the unknown number of i.i.d. segments of the environment, and C∈[0,T]C\in[0,T] being another unknown non-stationary measure defined as the minimal ℓ1\ell_1 deviation of the mean outcomes, our algorithms attain O~(min⁡{T+(TC)13,WT})\widetilde{\mathcal{O}}(\min\{\sqrt{T}+(TC)^{\frac{1}{3}}, \sqrt{WT}\}) for ℓ1\ell_1 calibration error and O~(min⁡{(1+C)13,W})\widetilde{\mathcal{O}}(\min\{(1+C)^{\frac{1}{3}}, W\}) for both ℓ2\ell_2 and pseudo KL calibration error. These bounds match the optimal rates in the stationary case (C=0C=0 and W=1W=1) and recover known guarantees in the fully adversarial regime (C,W=Ω(T)C, W=Ω(T)). Our approach builds on and extends prior work (Hu et al., 2026; Luo et al., 2025), introducing an epoch-based scheduling together with a novel non-uniform partition of the prediction space that allocates finer resolution near the underlying ground truth.

查看此版本解读 →查看此版本原文 ↗