A framework proves that broad recalibrated leakage is undetectable from predictions alone without an external discrimination ceiling, while near-label leaks produce a detectable unit-purity signature yielding a prior-free test.
arXiv:2108.02497
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2verdicts
UNVERDICTED 2representative citing papers
Proposes bearing-wise data partitioning to remove leakage in ML bearing fault diagnosis, reformulates as multi-label classification, and shows training bearing count drives generalization on four public datasets.
citing papers explorer
-
A prior-free blind detection of information leakage from model predictions
A framework proves that broad recalibrated leakage is undetectable from predictions alone without an external discrimination ceiling, while near-label leaks produce a detectable unit-purity signature yielding a prior-free test.