AWhat changed
Adds a finite-sample worst-case correction to held-out empirical loss and selects conservativeness to control expected loss on a future exchangeable example.
Scientific scope
Adds a finite-sample worst-case correction to held-out empirical loss and selects conservativeness to control expected loss on a future exchangeable example. Expectation is over calibration and test randomness, unlike high-probability calibration-conditional RCPS. Requires exchangeable, non-increasing, right-continuous losses with stated upper bound and feasible endpoint; arbitrary non-monotone losses are explicitly not covered. Near-tight lower bound has additional assumptions.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Expectation is over calibration and test randomness, unlike high-probability calibration-conditional RCPS. Requires exchangeable, non-increasing, right-continuous losses with stated upper bound and feasible endpoint; arbitrary non-monotone losses are explicitly not covered. Near-tight lower bound has additional assumptions.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction