mlsanityData to 14 Oct 2026

Accurate Inference for Adaptive Linear Models

arXiv 1712.06695 ↗First posted 18 Dec 2017stat.ML cs.LGcited by 39

Not on the map. Why?

Abstract (arXiv)

Estimators computed from adaptively collected data do not behave like their non-adaptive brethren. Rather, the sequential dependence of the collection policy can lead to severe distributional biases that persist even in the infinite data limit. We develop a general method -- W-decorrelation -- for transforming the bias of adaptive linear regression estimators into variance. The method uses only coarse-grained information about the data collection policy and does not need access to propensity scores or exact knowledge of the policy. We bound the finite-sample bias and variance of the W-estimator and develop asymptotically correct confidence intervals based on a novel martingale central limit theorem. We then demonstrate the empirical benefits of the generic W-decorrelation procedure in two different adaptive data settings: the multi-armed bandit and the autoregressive time series.

Key facts. Cited by 39 papers in the map (data through 14 Oct 2026). First posted on arXiv 18 Dec 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 13; 2022: 4; 2023: 5; 2024: 9; 2025: 6; 2026: 2.

39 papers.

Sort:

By area:

  1. 2021Statistical Inference with M-Estimators on Adaptively Collected Datacited by 40
  2. 2021Deeply-Debiased Off-Policy Interval Estimationcited by 36
  3. 2021A Closer Look at the Worst-case Behavior of Multi-armed Bandit Algorithmscited by 30
  4. 2021Near-optimal inference in adaptive linear regressioncited by 21
  5. 2024UCB algorithms for multi-armed bandits: Precise regret and adaptive inferencecited by 19
  6. 2021Metalearning Linear Bandits by Prior Updatecited by 16
  7. 2024Inference with the Upper Confidence Bound Algorithmcited by 12
  8. 2021From Finite to Countable-Armed Banditscited by 8
  9. 2022Synthetically Controlled Banditscited by 8
  10. 2022Online Statistical Inference for Contextual Bandits via Stochastic Gradient Descentcited by 8
  11. 2024Linear Contextual Bandits with Interferencecited by 8
  12. 2025Stable Thompson Sampling: Valid Inference via Variance Inflationcited by 8
27 more
  1. 2021Efficient Inference Without Trading-off Regret in Bandits: An Allocation Probability Test for Thompson Samplingcited by 7
  2. 2023Semi-parametric inference based on adaptively collected datacited by 7
  3. 2025Contextual Online Uncertainty-Aware Preference Learning for Human Feedbackcited by 7
  4. 2021Challenges in Statistical Analysis of Data Collected by a Bandit Algorithm: An Empirical Exploration in Applications to Adaptively Randomized Experimentscited by 6
  5. 2023Adaptive Linear Estimating Equationscited by 6
  6. 2024reBandit: Random Effects based Online RL algorithm for Reducing Cannabis Usecited by 6
  7. 2021Adaptive Data Debiasing through Bounded Explorationcited by 5
  8. 2023Adaptive Principal Component Regression with Applications to Panel Datacited by 5
  9. 2026Reinforcement Learning in the Real World: A Survey of Statistical Challenges and Future Directionscited by 5
  10. 2021Doubly Robust Interval Estimation for Optimal Policy Evaluation in Online Learningcited by 4
  11. 2023Online Estimation and Inference for Robust Policy Evaluation in Reinforcement Learningcited by 4
  12. 2024Best of Three Worlds: Adaptive Experimentation for Digital Marketing in Practicecited by 4
  13. 2025Adaptive Off-Policy Inference for M-Estimators Under Model Misspecificationcited by 4
  14. 2021Debiasing Samples from Online Learning Using Bootstrapcited by 3
  15. 2023Statistical Limits of Adaptive Linear Models: Low-Dimensional Estimation and Inferencecited by 3
  16. 2024Replicability is Asymptotically Free in Multi-armed Banditscited by 2
  17. 2024Oralytics Reinforcement Learning Algorithmcited by 2
  18. 2024MiWaves Reinforcement Learning Algorithmcited by 2
  19. 2025Avoiding the Price of Adaptivity: Inference in Linear Contextual Bandits via Stabilitycited by 2
  20. 2021Algorithms for Adaptive Experiments that Trade-off Statistical Analysis with Reward: Combining Uniform Random Assignment and Reward Maximizationcited by 1
  21. 2022Some performance considerations when using multi-armed bandit algorithms in the presence of missing datacited by 1
  22. 2022Entropy Regularization for Population Estimationcited by 1
  23. 2024Generalization Error Bounds for Learning under Censored Feedbackcited by 1
  24. 2021Statistical Consequences of Dueling Banditscited by 0
  25. 2025Adaptive Bounded Exploration and Intermediate Actions for Data Debiasingcited by 0
  26. 2025Bernstein-von Mises for Adaptively Collected Datacited by 0
  27. 2026Bandit Simulation for Average Reward Inferencecited by 0