Admissibility of Completely Randomized Trials: A Large-Deviation Approach

Chao Qin; Guido Imbens; Stefan Wager

REVIEW 3 major objections 5 minor 2 cited by

Admissibility of Completely Randomized Trials: A Large-Deviation Approach

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper proves that simple batched arm elimination designs strictly dominate completely randomized trials for best-arm identification whenever at least three arms are available.

desk verdict Elegant dominance result for BAI that hinges on an unproved joint LDP and a boundary example; send to review with a demand for the missing proof. read the letter →

arxiv 2506.05329 v1 pith:WCZOBPID submitted 2025-06-05 stat.ML cs.LGecon.EM

Guido Imbens , Chao Qin , Stefan Wager This is my paper

classification stat.MLcs.LGecon.EM MSC 60F1062L0562F07

keywords best-armidentificationbatchedarmeliminationcompletelyrandomizedtriallargedeviationsefficiencyexponentadmissibledesignsadaptiveexperimentsGaussianbandits

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether an experimenter who could run an adaptive trial is ever justified in running a non-adaptive completely randomized trial (CRT). It answers no for the best-arm identification problem when there are at least three treatment arms: there exist simple batched arm elimination designs whose large-sample error probability decays exponentially faster than that of a CRT on every Gaussian instance with known variance. The comparison is made through the efficiency exponent, the rate at which the chance of deploying the wrong arm vanishes as the sample grows. Establishing the dominance settles an open problem about the admissibility of completely randomized trials among adaptive designs.

What carries the argument

The central object is the batched arm elimination (BAE) design, a successive-rejects style rule that at pre-specified checkpoints discards the arm with the lowest empirical mean. Its error probability is controlled by two quantities: $w_n$, the total sample fraction accumulated toward the arm eliminated when $n$ arms remain, and $\Gamma_{\theta,n}$, the minimal Kullback-Leibler information needed for the best arm to look like the worst member of an $n$-arm set. The lower bound $e^{\mathrm{BAE}}_{\theta} \ge \min_n w_n \Gamma_{\theta,n}$, combined with Lemma 3's bound $\Gamma_{\theta,n} \ge \frac{n-1}{n}\frac{\Delta_{\min}(\theta)^2}{2\sigma^2}$, converts the design question into the algebraic condition $\min_n w_n(n-1)/n > 1/(2K)$. The exponential-rate proof itself rests on a joint large-deviation principle for empirical means and allocation sequences, stated as Theorem 8 and extended from bounded observations to Gaussian ones.

What would settle it

Simulate the two-batch design with $K=3$ arms, Gaussian outcomes with common variance $\sigma^2=1$ and means $(0,0,\Delta)$ for $\Delta>0$, and first-batch fraction $\beta_3=0.8$; estimate $-\frac{1}{T}\log P(\text{misidentification})$ at $T=10^4,10^5,10^6$. The paper's Theorem 5 predicts the limiting rate is strictly above the CRT rate $\Delta^2/12$; a rate converging to $\Delta^2/12$ or below would falsify the central claim.

Watch

Extended reading notes

Core claim

The paper's central claim is Theorem 5: under Gaussian arms with known common variance and a unique best arm, every batched arm elimination design whose batch weights satisfy $\min_{n=K,\ldots,2} w_n\frac{n-1}{n} > \frac{1}{2K}$ achieves $e^{\mathrm{BAE}}_{\theta} > e^{\mathrm{Unif}}_{\theta}$ for every instance $\theta$. Here $w_n$ is the fraction of the sample allocated, by the end of the batch that begins with $n$ arms, to the arm that is then eliminated, and the efficiency exponent $e^\pi_{\theta}$ measures the exponential decay rate of the misidentification probability, equivalently of the utilitarian regret. Since the CRT exponent is $\Delta_{\min}(\theta)^2/(4K\sigma^2)$ and the BAE lower bound is $\Delta_{\min}(\theta)^2/(2\sigma^2) \min_n w_n(n-1)/n$, the algebraic condition marks exactly where the adaptive design pulls ahead. The paper highlights a two-batch special case: randomize on all $K$ arms for a fraction $\beta_K$ of the sample, eliminate the $s$ worst arms, then randomize uniformly on the survivors, with $\beta_K$ large enough (for $s=1$, $\beta_K > 1/2 + 1/(2(K-1))$).

Load-bearing premise

The proof of the error-probability lower bound relies on a large-deviation theorem for Gaussian observations that is stated without proof in the appendix; if that theorem fails for adaptive, state-dependent allocation rules like BAE, the strict dominance result is unsupported.

Editorial extensions

If this is right

Whenever $K \ge 3$ and the batch weights satisfy $\min_n w_n(n-1)/n > 1/(2K)$, the error probability of BAE decays exponentially faster than that of complete randomization on every Gaussian instance with known variance, and the same ordering holds for utilitarian regret.
The two-batch rule with $s=1$ requires only that the first batch be larger than half the sample, namely $\beta_K > 1/2 + 1/(2(K-1))$, so the dominating design is simple to implement and needs no problem-specific tuning.
Completely randomized trials are not large-deviation admissible among fixed-budget adaptive policies when $K \ge 3$, settling the open question in the negative.
The sufficient condition spells out the design check: for each number of remaining arms, the per-arm cumulative allocation fraction, weighted by $(n-1)/n$, must clear the uniform-allocation threshold of $1/(2K)$.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

An experimenter reading these results would be tempted to adopt the two-batch rule in practice; the numerical example suggests the advantage survives a skewed, zero-inflated outcome distribution, but the stated theorem only covers Gaussian arms with known common variance.
Because the proof depends on an unproved Gaussian extension of a bounded-observation large-deviation theorem, the cleanest test of the paper's core claim is to verify Theorem 8 directly for the two-batch design rather than to re-simulate finite-sample regret.
The same weighted-allocation condition could serve as a template for other batched rules, such as eliminating several arms per batch or using unequal batch lengths, whenever the per-arm cumulative allocation stays above the threshold.
If the large-deviation extension holds, the admissibility question for other non-adaptive benchmarks, such as stratified or matched-pair designs, becomes a natural next target, since the argument here is built specifically around uniform randomization as the baseline.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The paper gives a clean sufficient condition for batched arm elimination (BAE) to strictly beat completely randomized trials in fixed-budget best-arm identification, and it claims to settle an open problem from Qin (2022). The main dominance theorem is genuinely new, the proof chain from Lemma 2 through Corollary 4 to Theorem 5 is transparent, and the algebra in Lemma 9 checks out. The batch weights are instance-agnostic and no fitting or circular reasoning enters. The citation pattern is honest, with the reliance on Wang et al. (2023) explicit.

The soft spots are real but concentrated. The entire lower bound on BAE's error exponent depends on Theorem 8, a joint large-deviation principle for non-anticipating algorithms with unbounded Gaussian observations. That theorem is stated without proof, described only as an extension of Wang et al.'s bounded-observation result. For a state-dependent adaptive allocation like BAE, this is not a cosmetic gap: one needs an LDP upper bound for the allocation sequence with a rate function that is never defined, plus a check that the rate is zero at the limiting allocation. No truncation argument is given. If Theorem 8 fails, the dominance result is unsupported. The numerical experiment does not test the LDP, only finite-sample regret, so it provides no reassurance on this point.

There is also a small but embarrassing boundary issue. The two-batch design showcased in the introduction and used in the numerics sets the first-batch fraction to K/(2(K-1)) (and to 2/3 for K=4), which is exactly the threshold in Example 1's condition, not strictly above it. The theorem requires a strict inequality, so the showcased design does not actually satisfy the paper's own sufficient condition. The fix is trivial—take the first batch slightly larger—but as written the example is outside the claimed result.

My overall take is conditional, not negative. The main idea is credible and the result would be important if Theorem 8 can be proved. I would send this to a serious referee rather than desk reject. The referee's primary job is to verify or refute Theorem 8; the authors should also fix the boundary example and either prove the LDP or cite a version that covers Gaussian adaptive allocations. This paper is worth a reading group discussion and, once the gap is closed, worth citing.

Referee Report

3 major / 5 minor

Summary. The paper studies best-arm identification with Gaussian arms of known variance and compares adaptive batched arm elimination (BAE) designs against completely randomized trials (CRTs) using the large-deviation efficiency exponent of the utilitarian regret. Under Assumption 1, the authors claim that whenever K >= 3, there exist simple instance-agnostic BAE designs whose efficiency exponent is strictly larger than that of a CRT for every problem instance (Theorem 5), resolving the second open problem posed in Qin [2022]. The proof combines a lower bound on the error exponent of general BAE designs (Lemma 2, via Lemma 7) with a lower bound on the minimal information quantity Gamma (Lemma 3). The key technical step is an unproved joint large-deviation principle stated as Theorem 8 in Appendix A.1, which is asserted as an extension of Wang et al. [2023] to Gaussian observations. A numerical experiment on a skewed charity-donation distribution is reported.

Significance. If the main claim holds, it is conceptually significant: it shows that an adaptive design can strictly dominate a non-adaptive benchmark instance-by-instance in a canonical bandit task, rather than only on average or in a minimax sense. The sufficient condition in Theorem 5 is explicit, and the paper identifies a very simple two-batch design. The algebraic work in Lemma 9 is correct, and the paper is generally well organized. The manuscript also contains a thoughtful discussion of related work and the distinction between fixed-budget and fixed-confidence settings. However, the central large-deviation theorem is stated without proof and is load-bearing for the main result, so the paper's conclusions are conditional on the validity of that theorem and on the verification of its hypotheses for BAE.

major comments (3)

[Appendix A.1, Theorem 8] Theorem 8 is the only bridge in Lemma 7 between the event {m_Tn in M, p_Tn in P} and the rate w_n Gamma_{theta,n}, but it is stated without proof. The claimed extension of Wang et al. [2023, Theorem 1] from bounded observations to unbounded Gaussian observations is nontrivial for an adaptive elimination algorithm, because the allocation can react to deviations of the empirical means and because unbounded rewards complicate change-of-measure arguments. The authors should either provide a complete proof of Theorem 8 or give a precise reference that covers the Gaussian case; otherwise Lemma 2 and Theorem 5 are unsupported.
[Lemma 7 and Theorem 8 hypotheses] To apply Theorem 8 to BAE, the paper must verify that the empirical allocation sequence {p_t} of BAE satisfies the assumed large-deviation upper bound with some rate function I, and it must define that rate function. The manuscript never does so. I note that the subsequent step of replacing inf_p max{F(p), I(p)} by inf_p F(p) in Lemma 7 is valid without any condition on I(p*), because max{F,I} >= F; the substantive gap is the unproved theorem and the unverified LDP condition on the allocation sequence.
[Section 4, numerical example] The simulation evaluates a design with K=4, s=1, and first batch fraction beta_K = 2/3. For these parameters, the sufficient condition in Example 1 requires beta_K > 1/2 + 1/(2(K-s)) = 1/2 + 1/6 = 2/3, so the simulated design is exactly at the boundary and is not covered by Theorem 5. The numerical experiment therefore cannot be cited as empirical support for the dominance result; the authors should either use a strictly larger first batch or report the simulation as an exploratory robustness check outside the proven range.

minor comments (5)

[Definition 2] The sentence 'a policy pi is large-deviation admissible is there is no policy' should read '... is admissible if there is no policy'.
[Algorithm 1, line 6] The variable r in the condition 'if r >= 2 and t = (beta_K + ... + beta_r)T' is not defined; it should be the current number of remaining arms n.
[Section 2] There is a typo: 'speical case' should be 'special case'.
[Section 4] The text refers to the design 'described in Corollary 1 with s = 1', but no Corollary 1 appears in the paper; the intended reference appears to be Example 1.
[Section 2] The notation ln(K) in the successive rejects weights is nonstandard and should be defined, e.g., as the K-th harmonic number over 2 plus terms.

Circularity Check

0 steps flagged · score 1.0 of 10

No circularity: the dominance result follows from analytic lower bounds and an explicit comparison, though the proof hinges on an unproved large-deviation extension (Theorem 8), which is a correctness gap rather than circularity.

full rationale

The derivation chain is not circular. The paper compares the exact CRT efficiency exponent (Proposition 1, from Russo 2020) with a lower bound for BAE (Lemma 2, via Lemma 7). The sufficient condition in Theorem 5 is an explicit algebraic comparison: min_n w_n (n-1)/n > 1/(2K) makes the BAE lower bound exceed the CRT exponent pointwise. No parameter is fitted to data, no post-hoc instance-dependent quantity is renamed as a prediction, and the batch weights are designer-chosen and instance-agnostic. The only self-referential element is the citation to Qin [2022] for the open problem being resolved; that citation frames the contribution but does not justify any mathematical step, so it is not load-bearing. The main caveat is that Lemma 7 invokes Theorem 8, a joint large-deviation principle stated without proof as an extension of Wang et al. [2023, Theorem 1] to unbounded Gaussian observations. The application also drops the rate function I(p) without explicitly verifying I(p)=0 on the relevant allocation set. This is a serious unproved step and a correctness risk, but it is not circular equivalence: Theorem 8 is external in origin and its assumptions are not defined in terms of the paper's conclusion. The numerical section does not validate the LDP, only finite-sample regret, but that again concerns support and robustness, not circularity. Overall, the central claim has independent mathematical content and does not reduce to its inputs by construction.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The central claim is a theorem under a Gaussian model with designer-chosen batch weights; the main unpaid input is the LDP in Theorem 8, which is asserted rather than derived. No new physical or statistical entities are postulated.

free parameters (1)

BAE batch weights (β_K, ..., β_2) = Not fitted; designer-chosen under constraint min_n w_n (n-1)/n > 1/(2K)
The construction is indexed by batch weights; the dominance theorem holds only for weights satisfying this explicit inequality. They are not estimated from data, so they do not constitute data fitting, but the result depends on this choice.

assumptions (3)

domain assumption Gaussian outcomes with common known variance: P(·|θ_i) = N(θ_i, σ^2) (Assumption 1).
Defines the instance class Theta and gives the KL divergence d(λ,θ) = (λ-θ)^2/(2σ^2) used throughout. The CRT exponent in Proposition 1 is inherited from this model.
ad hoc to paper Theorem 8: joint large-deviation lower bound for empirical means given an allocation LDP, extended to Gaussian observations.
Stated without proof in Appendix A.1; it is the load-bearing step in Lemma 6 and Lemma 7. The paper says it extends Wang et al. [2023, Theorem 1] but does not show the extension.
ad hoc to paper The BAE allocation sequence {p_t} satisfies the assumed large-deviation upper bound with some rate function I(p).
Needed to apply Theorem 8 to BAE; the rate function and its properties are not verified in the paper.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Admissibility of Completely Randomized Trials: A Large-Deviation Approach." pith.science (2026). https://pith.science/paper/WCZOBPID

@misc{pith2026250605329,
  author       = {Pith},
  title        = {Pith review of: Admissibility of Completely Randomized Trials: A Large-Deviation Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WCZOBPID}},
  note         = {Machine review of arXiv:2506.05329}
}

read the original abstract

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].

Figures

Figures reproduced from arXiv: 2506.05329 by the authors.

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Minimax and Bayes Optimal Best-Arm Identification
econ.EM 2025-06 conditional novelty 8.0 of 10

TS-SPAS attains the exact asymptotic minimax and Bayes constants for fixed-budget best-arm identification, with matching lower and upper bounds over exponential family outcomes.
Fundamental Limitations of Fixed-Budget Best-Arm Identification
cs.LG 2026-07 accept novelty 7.0 of 10

For K≥3 arms from any one-parameter natural exponential family, every algorithm falls short of the static oracle error-decay rate by a factor at least (1+log(K)/8) on some instance, so fixed-budget BAI admits no complexity.

Reference graph

Works this paper leans on

29 extracted references · 22 canonical work pages · cited by 2 Pith papers

[1]

A daptive treatment assignment in experiments for policy choice

Kaito Ariu, Masahiro Kato, Junpei Komiyama, Kenichiro McAlinn, and Chao Qin. A comment on " A daptive treatment assignment in experiments for policy choice". Econometrica, 2025

work page 2025
[2]

Best arm identification in multi-armed bandits

Jean-Yves Audibert, S \'e bastien Bubeck, and R \'e mi Munos. Best arm identification in multi-armed bandits. In COLT-23th Conference on learning theory-2010, pages 13--p, 2010

work page 2010
[3]

An adaptive targeted field experiment: Job search assistance for refugees in J ordan

A Stefano Caria, Grant Gordon, Maximilian Kasy, Simon Quinn, Soha Osman Shami, and Alexander Teytelboym. An adaptive targeted field experiment: Job search assistance for refugees in J ordan. Journal of the European Economic Association, 22 0 (2): 0 781--836, 2024

work page 2024
[4]

An empirical evaluation of T hompson sampling

Olivier Chapelle and Lihong Li. An empirical evaluation of T hompson sampling. In Advances in Neural Information Processing Systems, volume 24, pages 2249--2257, 2011

work page 2011
[5]

Sequential design of experiments

Herman Chernoff. Sequential design of experiments. The Annals of Mathematical Statistics, 30 0 (3): 0 755--770, 1959

work page 1959
[6]

On the existence of a complexity in fixed budget bandit identification

R \'e my Degenne. On the existence of a complexity in fixed budget bandit identification. In Proceedings of Thirty Sixth Conference on Learning Theory. PMLR, 2023

work page 2023
[7]

R. A. Fisher. The design of experiments. The Design of Experiments, 1925

work page 1925
[8]

Optimal best arm identification with fixed confidence

Aur \'e lien Garivier and Emilie Kaufmann. Optimal best arm identification with fixed confidence. In Conference on Learning Theory, pages 998--1027. PMLR, 2016

work page 2016

Show all 29 references

[9]

A large deviations perspective on ordinal optimization

Peter Glynn and Sandeep Juneja. A large deviations perspective on ordinal optimization. In Proceedings of the 2004 Winter Simulation Conference, 2004., volume 1. IEEE, 2004

2004
[10]

Confidence intervals for policy evaluation in adaptive experiments

Vitor Hadad, David A Hirshberg, Ruohan Zhan, Stefan Wager, and Susan Athey. Confidence intervals for policy evaluation in adaptive experiments. Proceedings of the National Academy of Sciences, 118 0 (15): 0 e2014602118, 2021

2021
[11]

Asymptotic representations for sequential decisions, adaptive experiments, and batched bandits

Keisuke Hirano and Jack R Porter. Asymptotic representations for sequential decisions, adaptive experiments, and batched bandits. arXiv preprint arXiv:2302.03117, 2023

2023 arXiv
[12]

Review on ranking and selection: A new perspective

L Jeff Hong, Weiwei Fan, and Jun Luo. Review on ranking and selection: A new perspective. Frontiers of Engineering Management, 8 0 (3): 0 321--343, 2021

2021
[13]

Causal inference in statistics, social, and biomedical sciences

Guido W Imbens and Donald B Rubin. Causal inference in statistics, social, and biomedical sciences. Cambridge University Press, 2015

2015
[14]

Does price matter in charitable giving? evidence from a large-scale natural field experiment

Dean Karlan and John A List. Does price matter in charitable giving? evidence from a large-scale natural field experiment. American Economic Review, 97 0 (5): 0 1774--1793, 2007

2007
[15]

Adaptive treatment assignment in experiments for policy choice

Maximilian Kasy and Anja Sautmann. Adaptive treatment assignment in experiments for policy choice. Econometrica, 89 0 (1): 0 113--132, 2021

2021
[16]

On the complexity of best-arm identification in multi-armed bandit models

Emilie Kaufmann, Olivier Capp \'e , and Aur \'e lien Garivier. On the complexity of best-arm identification in multi-armed bandit models. Journal of Machine Learning Research, 17 0 (1): 0 1--42, 2016

2016
[17]

Asymptotically efficient adaptive allocation rules

Tze Leung Lai and Herbert Robbins. Asymptotically efficient adaptive allocation rules. Advances in Applied Mathematics, 6 0 (1): 0 4--22, 1985

1985
[18]

Bandit algorithms

Tor Lattimore and Csaba Szepesv \'a ri. Bandit algorithms. Cambridge University Press, 2020

2020
[19]

Statistical inference for the mean outcome under a possibly non-unique optimal treatment strategy

Alexander R Luedtke and Mark J van der Laan. Statistical inference for the mean outcome under a possibly non-unique optimal treatment strategy. The Annals of Statistics, 44 0 (2): 0 713, 2016

2016
[20]

Batched bandit problems

Vianney Perchet, Philippe Rigollet, Sylvain Chassang, and Erik Snowberg. Batched bandit problems. The Annals of Statistics, 44 0 (2): 0 660--681, 2016

2016
[21]

Open problem: Optimal best arm identification with fixed-budget

Chao Qin. Open problem: Optimal best arm identification with fixed-budget. In Proceedings of Thirty Fifth Conference on Learning Theory, pages 5650--5654. PMLR, 02--05 Jul 2022

2022
[22]

Adaptive experimentation in the presence of exogenous nonstationary variation

Chao Qin and Daniel Russo. Adaptive experimentation in the presence of exogenous nonstationary variation. arXiv preprint arXiv:2202.09036, 2022

2022 arXiv
[23]

Optimizing adaptive experiments: A unified approach to regret minimization and best-arm identification

Chao Qin and Daniel Russo. Optimizing adaptive experiments: A unified approach to regret minimization and best-arm identification. arXiv preprint arXiv:2402.10592, 2024

2024 arXiv
[24]

Simple bayesian algorithms for best-arm identification

Daniel Russo. Simple bayesian algorithms for best-arm identification. Operations Research, 68 0 (6): 0 1625--1647, 2020

2020
[25]

Statistical decision functions

Abraham Wald. Statistical decision functions. The Annals of Mathematical Statistics, pages 165--205, 1949

1949
[26]

Best arm identification with fixed budget: A large deviation perspective

Po-An Wang, Ruo-Chun Tzeng, and Alexandre Proutiere. Best arm identification with fixed budget: A large deviation perspective. In Thirty-seventh Conference on Neural Information Processing Systems, 2023

2023
[27]

On universally optimal algorithms for A / B testing

Po-An Wang, Kaito Ariu, and Alexandre Proutiere. On universally optimal algorithms for A / B testing. In Proceedings of the 41st International Conference on Machine Learning, volume 235, pages 50065--50091, 2024

2024
[28]

Partial likelihood T hompson sampling

Han Wu and Stefan Wager. Partial likelihood T hompson sampling. In Uncertainty in Artificial Intelligence, pages 2138--2147. PMLR, 2022

2022
[29]

Inference for batched bandits

Kelly Zhang, Lucas Janson, and Susan Murphy. Inference for batched bandits. In Advances in Neural Information Processing Systems, volume 33, pages 9818--9829, 2020

2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

A daptive treatment assignment in experiments for policy choice

Kaito Ariu, Masahiro Kato, Junpei Komiyama, Kenichiro McAlinn, and Chao Qin. A comment on " A daptive treatment assignment in experiments for policy choice". Econometrica, 2025

work page 2025

[2] [2]

Best arm identification in multi-armed bandits

Jean-Yves Audibert, S \'e bastien Bubeck, and R \'e mi Munos. Best arm identification in multi-armed bandits. In COLT-23th Conference on learning theory-2010, pages 13--p, 2010

work page 2010

[3] [3]

An adaptive targeted field experiment: Job search assistance for refugees in J ordan

A Stefano Caria, Grant Gordon, Maximilian Kasy, Simon Quinn, Soha Osman Shami, and Alexander Teytelboym. An adaptive targeted field experiment: Job search assistance for refugees in J ordan. Journal of the European Economic Association, 22 0 (2): 0 781--836, 2024

work page 2024

[4] [4]

An empirical evaluation of T hompson sampling

Olivier Chapelle and Lihong Li. An empirical evaluation of T hompson sampling. In Advances in Neural Information Processing Systems, volume 24, pages 2249--2257, 2011

work page 2011

[5] [5]

Sequential design of experiments

Herman Chernoff. Sequential design of experiments. The Annals of Mathematical Statistics, 30 0 (3): 0 755--770, 1959

work page 1959

[6] [6]

On the existence of a complexity in fixed budget bandit identification

R \'e my Degenne. On the existence of a complexity in fixed budget bandit identification. In Proceedings of Thirty Sixth Conference on Learning Theory. PMLR, 2023

work page 2023

[7] [7]

R. A. Fisher. The design of experiments. The Design of Experiments, 1925

work page 1925

[8] [8]

Optimal best arm identification with fixed confidence

Aur \'e lien Garivier and Emilie Kaufmann. Optimal best arm identification with fixed confidence. In Conference on Learning Theory, pages 998--1027. PMLR, 2016

work page 2016