Cross-sample prediction churn between bootstrap-trained classifiers reaches 8-22% on chemistry benchmarks; K-bootstrap bagging reduces it 40-54% and twin-bootstrap with sym-KL consistency loss reduces it a further median 45% at matched 2x compute.
Churn reduction via distillation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Reducing cross-sample prediction churn in scientific machine learning
Cross-sample prediction churn between bootstrap-trained classifiers reaches 8-22% on chemistry benchmarks; K-bootstrap bagging reduces it 40-54% and twin-bootstrap with sym-KL consistency loss reduces it a further median 45% at matched 2x compute.