Research map / Multi-agent systems and games / Cooperative multi-agent RL
Multi-Agent Multi-Armed Bandits with Limited Communication
Mridul Agarwal, Vaneet Aggarwal, Kamyar Azizzadenesheli
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 LCC-UCB instead of no communication | LCC-UCB · Fully connected bandits (N=10, K=100; N=20, K=100; N=10, K=200) | cumulative regret | Improved |
| use LCC-UCB instead of DEMAB | LCC-UCB · Fully connected bandits (N=10, K=100; N=20, K=100; N=10, K=200) | cumulative regret | Improved |
| use DEMAB instead of no communication | Fully connected bandits · Fully connected bandits (N=10, K=100; N=20, K=100; N=10, K=200) | cumulative regret | Hurt |
| use LCC-UCB with limited communication instead of communication at every time step | LCC-UCB · Fully connected bandits (N=10, K=100; N=10, K=200) | cumulative regret | Mixed |
| use LCC-UCB-GRAPH instead of no communication | LCC-UCB-GRAPH · Sparse graphs (N=100, K=250; N=150, K=250; N=100, K=500) | cumulative regret | Improved |
| use LCC-UCB-GRAPH instead of communication after every time step with neighbors | LCC-UCB-GRAPH · Sparse graphs (N=100, K=250; N=150, K=250; N=100, K=500) | cumulative regret | Improved |
| use LCC-UCB-GRAPH instead of DEMAB | LCC-UCB-GRAPH · Sparse graphs (N=100, K=250; N=150, K=250; N=100, K=500) | cumulative regret | Improved |
| increase the number of arms from 250 to 500 | LCC-UCB-GRAPH · Sparse graphs (N=100) | cumulative regret growth rate | Improved |
| use LCC-UCB-GRAPH instead of relay-based LCC-UCB (ablation) | LCC-UCB-GRAPH · Sparse graphs | cumulative regret | Improved |
Closest papers by meaning
- Communication-Efficient Collaborative Regret Minimization in Multi-Armed Bandits (AAAI 2024 · similarity 0.88)
- Fair Algorithms for Multi-Agent Multi-Armed Bandits (NeurIPS 2021 · similarity 0.88)
- Collaborative Multi-Agent Heterogeneous Multi-Armed Bandits (ICML 2023 · similarity 0.87)
- Distributed Multi-Agent Bandits Over Erdős-Rényi Random Networks (NeurIPS 2025 · similarity 0.87)
- Heterogeneous Multi-Agent Bandits with Parsimonious Hints (AAAI 2025 · similarity 0.87)
- Individual Regret in Cooperative Stochastic Multi-Armed Bandits (NeurIPS 2025 · 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.
