Low-Rank MDPs with Continuous Action Spaces

arXiv:2311.03564v2 Announce Type: replace
Abstract: Low-Rank Markov Decision Processes (MDPs) have recently emerged as a promising framework within the domain of reinforcement learning (RL), as they allow for provably approximately correct (PAC) learning guarantees while also incorporating ML algorithms for representation learning. However, current methods for low-rank MDPs are limited in that they only consider finite action spaces, and give vacuous bounds as $|\mathcal{A}| \to \infty$, which greatly limits their applicability. In this work, we study the problem of extending …

abstract algorithms arxiv continuous cs.ai cs.lg current decision domain framework however low markov ml algorithms processes reinforcement reinforcement learning representation representation learning spaces stat.ml type