Bounded Aggregation for Continuous Time Markov Decision Processes.
Peter BuchholzIryna DohndorfAlexander FrankDimitri ScheftelowitschPublished in: EPEW (2017)
Keyphrases
- markov decision processes
- state space
- reinforcement learning
- finite state
- optimal policy
- stationary policies
- markov chain
- dynamic programming
- reachability analysis
- average cost
- policy iteration
- partially observable
- transition matrices
- decision processes
- reinforcement learning algorithms
- action space
- risk sensitive
- finite horizon
- dynamical systems
- infinite horizon
- factored mdps
- decision theoretic planning
- planning under uncertainty
- optimal control
- interval estimation
- model based reinforcement learning
- markov decision process
- reward function
- action sets
- learning algorithm
- stochastic shortest path