Rolling-origin conformal prediction achieves a minimax-optimal coverage error rate of O(T^{-β/(2β+1)}) under Hölder-β local stationarity and α-mixing by using a data-driven calibration window of size m* ≍ T^{2β/(2β+1)}.
For heavy-tailed oracle scores where |Yj| ≤ 1 fails, Rio’s Fuk–Nagaev inequality (Rio
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Rolling-Origin Conformal Prediction under Local Stationarity and Weak Dependence
Rolling-origin conformal prediction achieves a minimax-optimal coverage error rate of O(T^{-β/(2β+1)}) under Hölder-β local stationarity and α-mixing by using a data-driven calibration window of size m* ≍ T^{2β/(2β+1)}.