A mixed value and policy iteration method for stochastic control with universally measurable policies
Year of publication: |
2015
|
---|---|
Authors: | Yu, Huizhen ; Bertsekas, Dimitri P. |
Published in: |
Mathematics of operations research. - Catonsville, MD : INFORMS, ISSN 0364-765X, ZDB-ID 195683-8. - Vol. 40.2015, 4, p. 926-968
|
Subject: | discrete-time stochastic control | Borel spaces Markov decision process | total cost criteria | measurability | value iteration | policy iteration | convergence | Stochastischer Prozess | Stochastic process | Markov-Kette | Markov chain | Kontrolltheorie | Control theory | Entscheidung | Decision |
-
Markov Decision Processes and Stochastic Positional Games : Optimal Control on Complex Networks
Lozovanu, Dmitrii, (2024)
-
Solving average cost Markov decision processes by means of a two-phase time aggregation algorithm
Arruda, E. F., (2015)
-
Rectangular sets of probability measures
Shapiro, Alexander, (2016)
- More ...
-
Error bounds for approximations from projected linear equations
Yu, Huizhen, (2010)
-
Q-learning and policy iteration algorithms for stochastic shortest path problems
Yu, Huizhen, (2013)
-
On boundedness of Q-learning iterates for stochastic shortest path problems
Yu, Huizhen, (2013)
- More ...