Standard dynamic programming applied to time aggregated Markov decision processes.
Edilson F. ArrudaMarcelo D. FragosoPublished in: CDC (2009)
Keyphrases
- markov decision processes
- dynamic programming
- optimal policy
- state space
- finite state
- reinforcement learning
- transition matrices
- infinite horizon
- risk sensitive
- planning under uncertainty
- partially observable
- decision theoretic planning
- model based reinforcement learning
- reachability analysis
- policy iteration
- action space
- action sets
- multistage
- state and action spaces
- average cost
- learning algorithm
- decision processes
- linear program
- decision diagrams
- finite horizon
- state abstraction
- partially observable markov decision processes
- reinforcement learning algorithms
- function approximation
- decision problems
- semi markov decision processes
- heuristic search