A Version of the Euler Equation in Discounted Markov Decision Processes.
Hugo Cruz-SuárezGabriel Zacarías-EspinozaVíctor Vázquez-GuevaraPublished in: J. Appl. Math. (2012)
Keyphrases
- markov decision processes
- optimal policy
- differential equations
- state space
- dynamic programming
- finite state
- reinforcement learning
- infinite horizon
- average reward
- policy iteration
- reinforcement learning algorithms
- average cost
- planning under uncertainty
- transition matrices
- finite horizon
- decision processes
- model based reinforcement learning
- factored mdps
- state and action spaces
- action space
- reachability analysis
- decision problems
- risk sensitive
- multistage
- discounted reward