Derives limiting out-of-sample risk showing double descent under simple weighting and flat risk under strategic weighting in high-dimensional multimodel prediction, and proposes LaMA method.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Double Descent, Ensemble Emergence, and Large Model Averaging in High-Dimensional Multimodel Prediction
Derives limiting out-of-sample risk showing double descent under simple weighting and flat risk under strategic weighting in high-dimensional multimodel prediction, and proposes LaMA method.