Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Batched Thompson Sampling
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 all observed rewards rather than only cycle-boundary rewards (ablation) | Batched Thompson sampling · Batched multi-armed bandits | empirical performance | Helped |
| Set α=1.00001, giving one cycle per batch (ablation) | Batched Thompson sampling · Batched multi-armed bandits, T=10^5 | empirical regret | No clear effect |
| Increase α (ablation) | Batched Thompson sampling · Batched multi-armed bandits | regret | Hurt |
| Increase α from very small values toward 2 (ablation) | Batched Thompson sampling · Batched multi-armed bandits, T=10^5 | batch-count stability across reward distributions | Helped |
| Use α=1.00001 instead of larger α (ablation) | Batched Thompson sampling · Batched multi-armed bandits, T=10^5 | average number of batches | Mixed |
Closest papers by meaning
- Parallelizing Thompson Sampling (NeurIPS 2021 · similarity 0.92)
- Almost Optimal Anytime Algorithm for Batched Multi-Armed Bandits (ICML 2021 · similarity 0.91)
- Feel-Good Thompson Sampling for Contextual Bandits: a Markov Chain Monte Carlo Showdown (NeurIPS 2025 · similarity 0.89)
- Optimal Batched Linear Bandits (ICML 2024 · similarity 0.89)
- The Choice of Noninformative Priors for Thompson Sampling in Multiparameter Bandit Models (AAAI 2024 · similarity 0.89)
- Thompson Sampling For Combinatorial Bandits: Polynomial Regret and Mismatched Sampling Paradox (NeurIPS 2024 · similarity 0.89)
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.