Communication Dans Un Congrès Année : 2025

ε-Optimally Solving Two-Player Zero-Sum POSGs

Résumé

We present a novel framework for ε-optimally solving two-player zero-sum partially observable stochastic games (zs-POSGs). These games pose a major challenge due to the absence of a principled connection with dynamic programming (DP) techniques developed for two-player zero-sum stochastic games (zs-SGs). Prior attempts at transferring solution methods have lacked a lossless reduction-defined here as a transformation that preserves value functions, equilibrium strategies, and optimality structure-thereby limiting generalisation to ad hoc algorithms. This work introduces the first lossless reduction from zs-POSGs to transition-independent zs-SGs, enabling the principled application of a broad class of DP-based methods. We show empirically that point-based value iteration (PBVI) algorithms, applied via this reduction, produce ε-optimal strategies across a range of benchmark domains, consistently matching or outperforming existing state-of-the-art methods. Our results open a systematic pathway for algorithmic and theoretical transfer from SGs to partially observable settings.

Fichier principal
Vignette du fichier
neurips_2025.pdf (471.22 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05375482 , version 1 (20-11-2025)

Licence

Identifiants

  • HAL Id : hal-05375482 , version 1

Citer

Erwan Escudie, Matthia Sabatelli, Olivier Buffet, Jilles Dibangoye. ε-Optimally Solving Two-Player Zero-Sum POSGs. Annual Conference on Neural Information Processing Systems, Dec 2025, San Diego (Californie - EU) et Mexico City (Mexique), United States. ⟨hal-05375482⟩
613 Consultations
166 Téléchargements

Partager

  • More