Batched arm elimination designs with a sufficiently large first batch achieve a strictly higher large-deviation efficiency exponent than completely randomized trials for every Gaussian instance with K at least 3.
A daptive treatment assignment in experiments for policy choice
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
stat.ML 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Admissibility of Completely Randomized Trials: A Large-Deviation Approach
Batched arm elimination designs with a sufficiently large first batch achieve a strictly higher large-deviation efficiency exponent than completely randomized trials for every Gaussian instance with K at least 3.