Optimal homogeneous extreme predictors are non-extreme conditional quantiles of tilted distributions from the angular measure, with universally consistent peaks-over-threshold estimators.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.ST 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
On the optimal prediction of extreme events
Optimal homogeneous extreme predictors are non-extreme conditional quantiles of tilted distributions from the angular measure, with universally consistent peaks-over-threshold estimators.