Satisfiability Bounds for ω-Regular Properties in Bounded-Parameter Markov Decision Processes.
Maximilian WeiningerTobias MeggendorferJan KretínskýPublished in: CDC (2019)
Keyphrases
- markov decision processes
- optimal policy
- state space
- finite state
- reinforcement learning
- transition matrices
- planning under uncertainty
- policy iteration
- discounted reward
- factored mdps
- dynamic programming
- infinite horizon
- finite horizon
- model based reinforcement learning
- reachability analysis
- state and action spaces
- reinforcement learning algorithms
- decision theoretic planning
- decision processes
- upper bound
- action sets
- average cost
- action space
- partially observable
- lower bound
- average reward
- computational complexity
- markov decision process
- risk sensitive
- upper and lower bounds
- optimal solution
- constraint satisfaction problems
- semi markov decision processes
- markov chain
- data mining