Metraipolis › archive › dc_2d391dc4052c

Decentralized Multi-Player Q-Learning in Episodic Markov Decision Processes with Information Asymmetry — arXiv:2608.12753

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.12753
License (as archived)
CC0-1.0
Archived
2026-10-19T04:27:25.573Z
SHA-256 of the archived text
8c94b6e1f4328b89fb2e8b96135586fc545162ccf56d50a2ddcaab9047663be4
Stable URL of this page
https://metraipolis.com/documento/dc_2d391dc4052c

How to cite

Decentralized Multi-Player Q-Learning in Episodic Markov Decision Processes with Information Asymmetry — arXiv:2608.12753 — arXiv (en), https://arxiv.org/abs/2608.12753 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2d391dc4052c (Metraipolis archive dc_2d391dc4052c, sha256 8c94b6e1f4328b89fb2e8b96135586fc545162ccf56d50a2ddcaab9047663be4)
BibTeX
@misc{dc2d391dc4052c,
  title = {Decentralized Multi-Player Q-Learning in Episodic Markov Decision Processes with Information Asymmetry — arXiv:2608.12753},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.12753},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2d391dc4052c; sha256 8c94b6e1f4328b89fb2e8b96135586fc545162ccf56d50a2ddcaab9047663be4},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.12753. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We study decentralized multi-player reinforcement learning in episodic tabular Markov decision processes (MDPs) under three forms of information asymmetry: (A) unobserved actions with common rewards, (B) observed actions with independent rewards, and (C) unobserved actions with independent rewards. Players cannot communicate during learning but may agree on a protocol a priori. For Problems A and B we propose \texttt{mQ-learning} and \texttt{mQ-learning-intervals}, achieving $\tilde{O}(\sqrt{H^4 S A_{\text{joint}}\, T})$ regret, where $H$ is the horizon, $S$ the state count, $T = KH$ the total steps, and $A_{\text{joint}} = \prod_{i=1}^M |\mathcal{A}_i|$ the joint action space across $M$ players. For Problem C we give \texttt{mEXC} and \texttt{mEXC-Bellman}, two-phase explore-then-commit algorithms with regret $\tilde{O}(H (S A_{\text{joint}})^{1/3} T^{2/3})$. Against the centralized joint-action benchmark, decentralized learning under information asymmetry matches the single-agent Q-learning rate of \cite{jin2018q} up to logarithmic factors. Because $A_{\text{joint}}$ grows exponentially in $M$, the bounds are most meaningful for small $M$ or small per-player action sets.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.12753 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T04:20:35.825Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).