A regret-minimization algorithm for online conformal prediction under adversarial semi-bandit feedback achieves long-run coverage guarantees while controlling prediction set size.
Note that OCP-Unlock+ achieves the tightest bound as long as( √ K− √ C) √ TlnK >2o(T) −1T
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Online Conformal Prediction with Adversarial Semi-bandit Feedback via Regret Minimization
A regret-minimization algorithm for online conformal prediction under adversarial semi-bandit feedback achieves long-run coverage guarantees while controlling prediction set size.