Guido Imbens, Chao Qin, Stefan Wager
arXiv 5 Jun 2025 · Statistics — Machine Learning
arXiv:2506.05329 · PDF · DOI · OpenAlex · Extracted main text
When an experimenter has the option of running an adaptive trial, is it admissible to ignore this option and run a non-adaptive trial instead? We provide a negative answer to this question in the best-arm identification problem, where the experimenter aims to allocate measurement efforts judiciously to confidently deploy the most effective treatment arm. We find that, whenever there are at least three treatment arms, there exist simple adaptive designs that universally and strictly dominate non-adaptive completely randomized trials. This dominance is characterized by a notion called efficiency exponent, which quantifies a design's statistical efficiency when the experimental sample is large. Our analysis focuses on the class of batched arm elimination designs, which progressively eliminate underperforming arms at pre-specified batch intervals. We characterize simple sufficient conditions under which these designs universally and strictly dominate completely randomized trials. These results resolve the second open problem posed in Qin [2022].
appendix boundary found by appendix_command · 69% of the source is main text. Read the extracted text to check this.
The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.
| Reference | Intensity | Mentions | Sections | Main text | |
|---|---|---|---|---|---|
| 1 | Jean-Yves Audibert, Sébastien Bubeck, and Rémi Munos (2010) Best arm identification in multi-armed bandits | 1.000 | 7 | 3 | 100% |
| 2 | Chao Qin (2022) Open problem: Optimal best arm identification with fixed-budget self | 0.874 | 6 | 2 | 100% |
| 3 | Po-An Wang, Ruo-Chun Tzeng, and Alexandre Proutiere (2023) Best arm identification with fixed budget: A large deviation perspective | 0.794 | 6 | 4 | 50% |
| 4 | Dean Karlan and John A List (2007) Does price matter in charitable giving? evidence from a large-scale natural field experiment | 0.737 | 3 | 2 | 100% |
| 5 | Herman Chernoff (1959) Sequential design of experiments | 0.644 | 4 | 1 | 100% |
| 6 | Rémy Degenne (2023) On the existence of a complexity in fixed budget bandit identification | 0.644 | 4 | 1 | 100% |
| 7 | Daniel Russo (2020) Simple bayesian algorithms for best-arm identification | 0.644 | 4 | 1 | 100% |
| 8 | Po-An Wang, Kaito Ariu, and Alexandre Proutiere (2024) On universally optimal algorithms for A/B testing | 0.644 | 2 | 2 | 100% |
| 9 | Aurélien Garivier and Emilie Kaufmann (2016) Optimal best arm identification with fixed confidence | 0.585 | 3 | 1 | 100% |
| 10 | Peter Glynn and Sandeep Juneja (2004) A large deviations perspective on ordinal optimization | 0.585 | 3 | 1 | 100% |
Showing the top 10 of 29 scored citations.