Partially Observable Stochastic Games with Neural Perception Mechanisms

arXiv:2310.11566v2 Announce Type: replace-cross
Abstract: Stochastic games are a well established model for multi-agent sequential decision making under uncertainty. In practical applications, though, agents often have only partial observability of their environment. Furthermore, agents increasingly perceive their environment using data-driven approaches such as neural networks trained on continuous data. We propose the model of neuro-symbolic partially-observable stochastic games (NS-POSGs), a variant of continuous-space concurrent stochastic games that explicitly incorporates neural perception mechanisms. We focus on a one-sided setting with a …

abstract agent agents applications arxiv continuous cs.gt cs.lg data data-driven decision decision making environment games making multi-agent networks neural networks observability observable perception practical stochastic type uncertainty