The Complexity of Graph-Based Reductions for Reachability in Markov Decision Processes.
Stéphane Le RouxGuillermo A. PérezPublished in: CoRR (2017)
Keyphrases
- markov decision processes
- state space
- finite state
- optimal policy
- reinforcement learning
- dynamic programming
- policy iteration
- transition matrices
- partially observable
- reinforcement learning algorithms
- decision theoretic planning
- average cost
- finite horizon
- action space
- model based reinforcement learning
- planning under uncertainty
- infinite horizon
- markov decision process
- decision processes
- computational complexity
- factored mdps
- risk sensitive
- reachability analysis
- average reward
- decision problems
- heuristic search
- action sets
- markov chain
- semi supervised
- partially observable markov decision processes
- continuous state spaces
- dynamical systems