Research map / Bandits and online learning / Multi-armed bandits
Efficient-UCBV: An Almost Optimal Algorithm Using Variance Estimates
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 |
|---|---|---|---|
| Combine variance-based confidence bounds with arm elimination | EUCBV · Stochastic MAB, gap-dependent regret | expected cumulative regret upper bound: EUCBV: $O\left( \dfrac{K\sigma_{\max}^{2}\log (\frac{T\Delta^2}{K})}{\Delta}\right)$; MOSS: $O\left( \dfrac{K^2\log (T\D | Helped |
| Use EUCBV's round-based elimination and variance-aware confidence bounds | EUCBV · Stochastic MAB, gap-independent regret | expected cumulative regret upper bound: EUCBV: $O\left(\sqrt{KT}\right)$; UCB-Improved: $O\left(\sqrt{KT\log K}\right)$ | Helped |
| Use variance-aware bounds and the proposed exploration parameters instead of UCBV's strategy | EUCBV relative to UCBV · Bernoulli with uniform gaps | expected cumulative regret | Helped |
| Allocate pulls according to arm variances | EUCBV · Gaussian 3 Group Mean Setting | expected cumulative regret | Helped |
| Use variance-aware exploration with the proposed exploration factors | EUCBV · Failure of TS | expected cumulative regret | Helped |
| Use variance-aware arm selection and elimination | EUCBV · Gaussian 3 Group Variance setting | expected cumulative regret | Helped |
| Use aggressive exploration parameters | EUCBV relative to UCBV · Experiments 1–4 | expected cumulative regret | Helped |
| Use non-uniform optimistic-greedy arm selection rather than equal pulls per arm each round | EUCBV · Stochastic MAB | expected cumulative regret | Helped |
Closest papers by meaning
- Precise Asymptotics and Refined Regret of Variance-Aware UCB (NeurIPS 2025 · similarity 0.94)
- A Closer Look at the Worst-case Behavior of Multi-armed Bandit Algorithms (NeurIPS 2021 · similarity 0.91)
- Best Arm Identification for Stochastic Rising Bandits (ICML 2024 · similarity 0.87)
- Only Pay for What Is Uncertain: Variance-Adaptive Thompson Sampling (ICLR 2024 · similarity 0.87)
- On the Asymptotic Optimality of Confidence Interval Based Algorithms for Fixed Confidence MABs (AAAI 2025 · similarity 0.87)
- Adaptive Regret for Bandits Made Possible: Two Queries Suffice (ICLR 2024 · similarity 0.87)
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.