Research map / Bandits and online learning / Multi-armed bandits
Multi-Player Bandits – a Musical Chairs Approach
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 random exploration followed by fixing each player on a distinct top-N arm | Static multi-player bandits · Static multi-player bandits | regret: T_0 \cdot N + 2 \cdot \exp(2) \cdot N^2 versus O\left(T^\frac{2}{3}\right) | Helped |
| Run learning and allocation in recurring epochs to adapt to player changes | Dynamic multi-player bandits · Dynamic multi-player bandits | regret: \tilde{O} \left( \sqrt{xT}\right) | Helped |
| Use MEGA under late entry followed by a player's exit | Dynamic multi-player bandits · Player enters at ⌈T/2⌉ and first player later leaves | regret: \Omega \left(T \right) | Hurt |
| Use DMC instead of MEGA under late entry followed by a player's exit | Dynamic multi-player bandits · Player enters at ⌈T/2⌉ and first player later leaves | regret: \tilde{O} \left(\sqrt{T} \right) | Helped |
| Use DMC rather than MEGA for alternating player entry and exit | Dynamic multi-player bandits · Alternating entry and exit every ⌈T^λ⌉ rounds, λ > β | regret upper bound: O \left( T^{1 - \left( \lambda - \beta \right)} \right) versus \tilde{O} \left( T^{1- \frac{\lambda}{2}} \right) | Helped |
| Use MC's fixed-arm phase rather than continued epsilon-greedy exploration | Static multi-player bandits · Static experiment | accumulated regret | Helped |
| Use epoch-based DMC in the moderate-horizon dynamic setting | Dynamic multi-player bandits · Dynamic experiments with smaller T | regret | Hurt |
| Use epoch-based DMC when the horizon is sufficiently large | Dynamic multi-player bandits · Dynamic experiments with sufficiently large T | regret | Helped |
Closest papers by meaning
- SIC-MMAB: Synchronisation Involves Communication in Multiplayer Multi-Armed Bandits (NeurIPS 2019 · similarity 0.95)
- Decentralized Stochastic Multi-Player Multi-Armed Walking Bandits (AAAI 2023 · similarity 0.92)
- Individual Regret in Cooperative Stochastic Multi-Armed Bandits (NeurIPS 2025 · similarity 0.89)
- Nash Social Welfare for Multi Armed Bandits: Trajectory-wise Expected and High Probability Regret (NeurIPS 2026 · similarity 0.88)
- Strategic Multi-Armed Bandit Problems Under Debt-Free Reporting (NeurIPS 2024 · similarity 0.88)
- Distributed Multi-Agent Bandits Over Erdős-Rényi Random Networks (NeurIPS 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.