Partially Observed Optimal Stochastic Control: Regularity, Optimality, Approximations, and Learning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Kara, Ali Devran, Yuksel, Serdar
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929653475704832
author Kara, Ali Devran
Yuksel, Serdar
author_facet Kara, Ali Devran
Yuksel, Serdar
contents In this review/tutorial article, we present recent progress on optimal control of partially observed Markov Decision Processes (POMDPs). We first present regularity and continuity conditions for POMDPs and their belief-MDP reductions, where these constitute weak Feller and Wasserstein regularity and controlled filter stability. These are then utilized to arrive at existence results on optimal policies for both discounted and average cost problems, and regularity of value functions. Then, we study rigorous approximation results involving quantization based finite model approximations as well as finite window approximations under controlled filter stability. Finally, we present several recent reinforcement learning theoretic results which rigorously establish convergence to near optimality under both criteria.
format Preprint
id arxiv_https___arxiv_org_abs_2412_06735
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Partially Observed Optimal Stochastic Control: Regularity, Optimality, Approximations, and Learning
Kara, Ali Devran
Yuksel, Serdar
Optimization and Control
Systems and Control
In this review/tutorial article, we present recent progress on optimal control of partially observed Markov Decision Processes (POMDPs). We first present regularity and continuity conditions for POMDPs and their belief-MDP reductions, where these constitute weak Feller and Wasserstein regularity and controlled filter stability. These are then utilized to arrive at existence results on optimal policies for both discounted and average cost problems, and regularity of value functions. Then, we study rigorous approximation results involving quantization based finite model approximations as well as finite window approximations under controlled filter stability. Finally, we present several recent reinforcement learning theoretic results which rigorously establish convergence to near optimality under both criteria.
title Partially Observed Optimal Stochastic Control: Regularity, Optimality, Approximations, and Learning
topic Optimization and Control
Systems and Control
url https://arxiv.org/abs/2412.06735