Research map / Bandits and online learning / Multi-armed bandits
Robust Outlier Arm Identification
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 a median-and-MAD outlier threshold instead of a mean-and-standard-deviation threshold | ROAI · Contaminated arm means | threshold robustness | Helped |
| Use ROAILUCB instead of RR/WRR | Outlier arm identification · Synthetic data | samples required at the same error rate: about 5x fewer samples | Helped |
| Use ROAIElim instead of random sampling | Outlier arm identification · Synthetic data, anytime performance | empirical stopping time | Helped |
| Use ROAIElim instead of random sampling | Outlier arm identification · Synthetic data, anytime performance | anytime error rate | No clear effect |
| Use RR/WRR instead of random sampling with the median-MAD threshold | Outlier arm identification · Synthetic data | anytime error rate | Hurt |
| Use ROAILUCB instead of the other evaluated algorithms | Outlier arm identification · Wine Quality dataset | anytime error rate | Helped |
| Use RR/WRR instead of random sampling | Outlier arm identification · Wine Quality dataset | anytime error rate | Hurt |
| Construct the threshold from a subset of arms rather than all arms | ROAI with known contamination upper bound · ROAI | sample complexity | Mixed |
Closest papers by meaning
- Identifying Outlier Arms in Multi-Armed Bandit (NeurIPS 2017 · similarity 0.90)
- Adaptive Double-Exploration Tradeoff for Outlier Detection (AAAI 2020 · similarity 0.88)
- PAC Identification of a Bandit Arm Relative to a Reward Quantile (AAAI 2017 · similarity 0.88)
- Balancing Performance and Costs in Best Arm Identification (NeurIPS 2025 · similarity 0.86)
- Almost Minimax Optimal Best Arm Identification in Piecewise Stationary Linear Bandits (NeurIPS 2024 · similarity 0.85)
- Fixed Confidence Best Arm Identification in the Bayesian Setting (NeurIPS 2024 · similarity 0.85)
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.