Distributed Policy Iteration for Scalable Approximation of Cooperative Multi-Agent Policies

@inproceedings{ phanAAMAS19,
    author      = "Thomy Phan and Kyrill Schmid and Lenz Belzner and Thomas Gabor and Sebastian Feld and Claudia Linnhoff-Popien",
    title       = "Distributed Policy Iteration for Scalable Approximation of Cooperative Multi-Agent Policies",
    year        = "2019",
    
    
    abstract    = "We propose Strong Emergent Policy (STEP) approximation, a scalable approach to learn strong decentralized policies for cooperative MAS with a distributed variant of policy iteration. For that, we use function approximation to learn from action recommendations of a decentralized multi-agent planning algorithm. STEP combines decentralized multi-agent planning with centralized learning, only requiring a generative model for distributed black box optimization. We experimentally evaluate STEP in two challenging and stochastic domains with large state and joint action spaces and show that STEP is able to learn stronger policies than standard multi-agent reinforcement learning algorithms, when combining multi-agent open-loop planning with centralized function approximation. The learned policies can be reintegrated into the multi-agent planning process to further improve performance.",
    url         = "https://arxiv.org/pdf/1901.08761.pdf",
    eprint      = "https://thomyphan.github.io/files/2019-aamas-ea.pdf",
    location    = "Montreal QC, Canada",
    publisher   = "International Foundation for Autonomous Agents and Multiagent Systems",
    booktitle   = "Extended Abstracts of the 18th International Conference on Autonomous Agents and MultiAgent Systems",
    pages       = "2162--2164",
    keywords    = "multi-agent learning, policy iteration, multi-agent planning",
    doi         = "https://dl.acm.org/doi/10.5555/3306127.3332044"
}