mlsanityData to 14 Oct 2026

Efficient Adaptive Experimentation with Noncompliance

arXiv 2505.17468 ↗First posted 23 May 2025stat.ME cs.LG stat.MLcited by 2

Not on the map. Why?

Abstract (arXiv)

We study the problem of estimating the average treatment effect (ATE) in adaptive experiments where treatment can only be encouraged -- rather than directly assigned -- via a binary instrumental variable. Building on semiparametric efficiency theory, we derive the efficiency bound for ATE estimation under arbitrary, history-dependent instrument-assignment policies, and show it is minimized by a variance-aware allocation rule that balances outcome noise and compliance variability. Leveraging this insight, we introduce AMRIV -- an Adaptive, Multiply-Robust estimator for Instrumental-Variable settings with variance-optimal assignment. AMRIV pairs (i) an online policy that adaptively approximates the optimal allocation with (ii) a sequential, influence-function-based estimator that attains the semiparametric efficiency bound while retaining multiply-robust consistency. We establish asymptotic normality, explicit convergence rates, and anytime-valid asymptotic confidence sequences that enable sequential inference. Finally, we demonstrate the practical effectiveness of our approach through empirical studies, showing that adaptive instrument assignment, when combined with the AMRIV estimator, yields improved efficiency and robustness compared to existing baselines.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 33 of its 65 references that are in the map. First posted on arXiv 23 May 2025.

Builds on

33 of 65 references.

References · 33

  1. 2015Estimation and Inference of Heterogeneous Treatment Effects using Random Forestscited by 600
  2. 2016Error bounds for approximations with deep ReLU networkscited by 574
  3. 2017Nonparametric regression using deep neural networks with ReLU activation functioncited by 433
  4. 2017Size-Independent Sample Complexity of Neural Networkscited by 290
  5. 2022Game-theoretic statistics and safe anytime-valid inferencecited by 155
  6. 2019Kernel Instrumental Variable Regressioncited by 118
  7. 2019Deep Generalized Method of Moments for Instrumental Variable Analysiscited by 88
  8. 2014Consistency of random forestscited by 85
  9. 2021Statistical Inference with M-Estimators on Adaptively Collected Datacited by 40
  10. 2022Anytime-valid off-policy inference for contextual banditscited by 37
  11. 2019Machine Learning Estimation of Heterogeneous Treatment Effects with Instrumentscited by 29
  12. 2021Time-uniform central limit theory and asymptotic confidence sequencescited by 29
21 more
  1. 2014Mondrian Forests: Efficient Online Random Forestscited by 25
  2. 2016Bounded, efficient and multiply robust estimation of average treatment effects using instrumental variablescited by 23
  3. 2020Efficient Adaptive Experimental Design for Average Treatment Effect Estimationcited by 23
  4. 2023Semiparametric Efficient Inference in Adaptive Experimentscited by 23
  5. 2022Estimating individual treatment effects under unobserved confounding using binary instrumentscited by 18
  6. 2023Clip-OGD: An Experimental Design for Adaptive Neyman Allocation in Sequential Experimentscited by 16
  7. 2024Active Adaptive Experimental Design for Treatment Effect Estimation with Covariate Choicescited by 12
  8. 2021Efficient Online Estimation of Causal Effects by Deciding What to Observecited by 11
  9. 2023Stochastic Online Instrumental Variable Regression: Regrets for Endogeneity and Bandit Feedbackcited by 7
  10. 2023When is the estimated propensity score better? High-dimensional analysis and bias correctioncited by 7
  11. 2023Adaptive Instrument Design for Indirect Experimentscited by 6
  12. 2024Adaptive Experimentation When You Can't Experimentcited by 6
  13. 2020Cross-Fitting and Averaging for Machine Learning Estimation of Heterogeneous Treatment Effectscited by 5
  14. 2024Anytime-Valid Inference for Double/Debiased Machine Learning of Causal Parameterscited by 5
  15. 2024Optimal Adaptive Experimental Design for Estimating Treatment Effectcited by 5
  16. 2024Targeted Sequential Indirect Experiment Designcited by 4
  17. 2024Logarithmic Neyman Regret for Adaptive Estimation of the Average Treatment Effectcited by 4
  18. 2024Nonparametric identification and efficient estimation of causal effects with instrumental variablescited by 3
  19. 2024Off-policy estimation with adaptively collected data: the power of online learningcited by 3
  20. 2024Estimating Heterogeneous Treatment Effects by Combining Weak Instruments and Observational Datacited by 2
  21. 2025Optimistic Algorithms for Adaptive Estimation of the Average Treatment Effectcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026What Do We Care About in Bandits with Noncompliance? BRACE: Bandits with Recommendations, Abstention, and Certified Effectscited by 0
  2. 2026Adaptive Experimentation for Censored Survival Outcomescited by 0