Summary of Federated Combinatorial Multi-agent Multi-armed Bandits, by Fares Fourati et al.
Federated Combinatorial Multi-Agent Multi-Armed Bandits
by Fares Fourati, Mohamed-Slim Alouini, Vaneet Aggarwal
First submitted to arxiv on: 9 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Discrete Mathematics (cs.DM); Multiagent Systems (cs.MA); Machine Learning (stat.ML)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The introduced federated learning framework is designed specifically for online combinatorial optimization with bandit feedback. It enables agents to select subsets of arms, observe noisy rewards without individual arm information, and cooperate/share information at specific intervals. The framework transforms offline resilient single-agent algorithms into online multi-agent algorithms, achieving an alpha-regret of no more than O(m^(-1/3+β) ψ^(1/(3+β)) T^((2+β)/(3+β))). This approach eliminates the ε approximation error and ensures sublinear growth with respect to time horizon T. It also demonstrates a linear speedup with increasing numbers of communicating agents, while being communication-efficient with only O(ψT^((β)/(β+1))) communication rounds. The framework has been successfully applied to online stochastic submodular maximization using various offline algorithms, yielding the first results for single-agent and multi-agent settings. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary In simple terms, this research paper introduces a new way for many agents to work together and make decisions online. They can choose which “arms” (options) to take based on feedback they receive, without knowing detailed information about each arm. The framework helps convert offline algorithms into online ones, ensuring the agents’ decisions are reasonable and don’t waste time. This approach is efficient in terms of communication between agents and has been tested successfully for various decision-making tasks. |
Keywords
» Artificial intelligence » Federated learning » Optimization
