Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Inference for Batched Bandits
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use Batched OLS (BOLS) instead of AW-AIPW | two-arm bandit inference · stationary, nT<=300 | Type-1 error control | Helped |
| use AW-AIPW instead of BOLS | two-arm bandit inference · stationary, nT>=500 | power | Helped |
| use BOLS instead of other estimators | two-arm bandit inference · non-stationary baseline reward | Type-1 error and power | Helped |
| use plain OLS normal approximation instead of batched estimator | OLS · bandit data with no unique optimal arm | Type-1 error | Hurt |
| use W-decorrelated estimator instead of BOLS | W-decorrelated · stationary | power | Hurt |
| use self-normalized martingale high-probability bound | hypothesis testing · stationary | power | Hurt |
Closest papers by meaning
- Simulation-Based Inference for Adaptive Experiments (NeurIPS 2025 · similarity 0.89)
- Robustly Improving Bandit Algorithms with Confounded and Selection Biased Offline Data: A Causal Approach (AAAI 2024 · similarity 0.88)
- Balancing Performance and Costs in Best Arm Identification (NeurIPS 2025 · similarity 0.87)
- Noise-Adaptive Confidence Sets for Linear Bandits and Application to Bayesian Optimization (ICML 2024 · similarity 0.87)
- Optimal and Practical Batched Linear Bandit Algorithm (ICML 2025 · similarity 0.87)
- Breaking the $\log(1/\Delta_2)$ Barrier: Better Batched Best Arm Identification with Adaptive Grids (ICLR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.