Semi-supervised estimators for total and explained treatment heterogeneity, with an optimally weighted procedure that is guaranteed to be at least as efficient as supervised estimation.
(S37) It follows that ( ˆQ(k) )2 −θ2 ETH = ( ˆQ(k) +θETH)( ˆQ(k) −θETH) (i) =Op(1)op(1) = op(1), (S38) where (i) holds from ( S36) and ( S37)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Semi-supervised inference for treatment heterogeneity
Semi-supervised estimators for total and explained treatment heterogeneity, with an optimally weighted procedure that is guaranteed to be at least as efficient as supervised estimation.