Research map / Bandits and online learning / Multi-armed bandits
Optimal Best-Arm Identification Methods for Tail-Risk Measures
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 empirical-likelihood KL-projection confidence intervals instead of truncation-based empirical-estimator intervals | CVaR confidence intervals for heavy-tailed distributions · Distributions in the moment-bounded class L | confidence-interval width | Helped |
| Update oracle sampling weights only at geometrically spaced times | Proposed CVaR best-arm identification algorithm · General arms with bounded raw (1+epsilon)-moment | computational cost | Helped |
| Update oracle sampling weights only at geometrically spaced times | Proposed CVaR best-arm identification algorithm · General arms with bounded raw (1+epsilon)-moment | sample complexity | Hurt |
| Use the proposed plug-in algorithm with empirical distributions and optimal allocation weights | CVaR best-arm identification · Three-arm Fisher-Tippett/Pareto mixture | sample complexity relative to lower bound | Helped |
| Increase the number of arms | Proposed tail-risk best-arm identification algorithm · Additional experiments | average stopping time | Hurt |
| Change the moment-bound parameter B | Proposed tail-risk best-arm identification algorithm · Additional experiments | sample complexity | Mixed |
Closest papers by meaning
- Concentration bounds for CVaR estimation: The cases of light-tailed and heavy-tailed distributions (ICML 2020 · similarity 0.92)
- Balancing Performance and Costs in Best Arm Identification (NeurIPS 2025 · similarity 0.88)
- Does 1/2-Tsallis-INF Also Work Well for Best-Arm Identification? (NeurIPS 2026 · similarity 0.87)
- Sharp Non-Asymptotic Analysis of the Penalized Challenger in $\beta$-EB-TCI for Bernoulli Bandits (NeurIPS 2026 · similarity 0.86)
- Almost Minimax Optimal Best Arm Identification in Piecewise Stationary Linear Bandits (NeurIPS 2024 · similarity 0.86)
- Fixing the Loose Brake: Exponential-Tailed Stopping Time in Best Arm Identification (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.