Research map / Bandits and online learning / Multi-armed bandits
On Explore-Then-Commit strategies
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 |
|---|---|---|---|
| Replace fixed-design exploration with SPRT-based sequential stopping in ETC (known gap) | ETC strategy, two-armed Gaussian bandit · two-armed Gaussian bandit, known gap | asymptotic regret constant (R/log T times Delta): factor of 4 reduction (4/Delta to 1/Delta) | Helped |
| Use fully sequential UCB-inspired Delta-UCB instead of ETC (known gap) | fully sequential strategy vs ETC · two-armed Gaussian bandit, known gap | asymptotic regret constant (R/log T times Delta): factor of 2 over SPRT-ETC (1/Delta to 1/(2Delta)) | Helped |
| Use BAI-based ETC with log(T/2n) threshold instead of fixed-design (unknown gap) | ETC strategy, unknown gap · two-armed Gaussian bandit, unknown gap | asymptotic regret constant (R/log T times Delta) | No clear effect |
| Use UCB* with smaller confidence bonus instead of UCB (unknown gap) | UCB · two-armed Gaussian bandit, unknown gap | empirical regret vs UCB | Helped |
| Fully sequential strategies vs best ETC strategies | known and unknown gap settings · two-armed Gaussian bandit | asymptotic regret factor relative to best ETC: factor of 2 | Helped |
| Use ETC with Track-and-Stop exploration instead of uniform/fixed exploration for K arms | ETC strategy, multi-armed Gaussian bandit · K arms with equal suboptimal means | asymptotic regret vs Lai-Robbins bound | Mixed |
Closest papers by meaning
- Forced Exploration in Bandit Problems (AAAI 2024 · similarity 0.90)
- Greedy Algorithms for Structured Bandits: A Sharp Characterization of Asymptotic Success / Failure (NeurIPS 2025 · similarity 0.90)
- Causal Bandits: The Pareto Optimal Frontier of Adaptivity, a Reduction to Linear Bandits, and Limitations around Unknown Marginals (ICML 2024 · similarity 0.89)
- Bayesian Regret Minimization in Offline Bandits (ICML 2024 · similarity 0.89)
- Tracking Most Significant Shifts in Infinite-Armed Bandits (ICML 2025 · similarity 0.89)
- Improved Regret Analysis in Gaussian Process Bandits: Optimality for Noiseless Reward, RKHS norm, and Non-Stationary Variance (ICML 2025 · similarity 0.88)
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.