Research map / Bandits and online learning / Multi-armed bandits
Active Learning with Safety Constraints
Romain Camilleri, Andrew Wagenmaker, Jamie Heather Morgenstern, Lalit K Jain, Kevin Jamieson
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 adaptive experimental design that balances safety-gap and optimality-gap estimation | Beside · Multi-Armed Bandit | sample complexity: up to about two times | Improved |
| Use adaptive experimental design that balances safety and optimality estimation | Beside · Linear Response Model, random instance | sample complexity | Improved |
| Combine safety-focused and difference-focused designs rather than using only XYdiff | Beside · Linear Response Model, hard instance | sample complexity | Improved |
| Use randomized active exploration and safety/difference-focused sampling instead of passive uniform sampling | Constrained classification · Adult | FDR | Improved |
| Use randomized active exploration and safety/difference-focused sampling instead of passive uniform sampling | Constrained classification · Adult | accuracy | Improved |
| Use active sampling rather than passive uniform sampling | Constrained classification · Adult | labels needed for a given FDR: half the number of samples | Improved |
| Use randomized active exploration and safety/difference-focused sampling instead of passive uniform sampling | Constrained classification · German Credit | TPR | Improved |
| Use randomized active exploration and safety/difference-focused sampling instead of passive uniform sampling | Constrained classification · German Credit | precision | Improved |
| Use active sampling rather than passive uniform sampling | Constrained classification · German Credit | labels needed for a given precision: half the number of samples | Improved |
Closest papers by meaning
- Best Arm Identification in Generalized Linear Bandits via Hybrid Feedback (NeurIPS 2026 · similarity 0.87)
- Balancing Performance and Costs in Best Arm Identification (NeurIPS 2025 · similarity 0.87)
- Feasible Action Search for Bandit Linear Programs via Thompson Sampling (ICML 2025 · similarity 0.87)
- Online Learning with Unknown Constraints (ICML 2025 · similarity 0.87)
- Almost Minimax Optimal Best Arm Identification in Piecewise Stationary Linear Bandits (NeurIPS 2024 · similarity 0.86)
- Learning Safe Control via On-the-Fly Bandit Exploration (ICML 2025 · similarity 0.86)
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.
