Strategy Synthesis for Markov Decision Processes and Branching-Time Logics.
Tomás BrázdilVojtech ForejtPublished in: CONCUR (2007)
Keyphrases
- markov decision processes
- state space
- optimal policy
- reachability analysis
- transition matrices
- finite state
- epistemic logic
- dynamic programming
- model checking
- reinforcement learning
- policy iteration
- finite horizon
- reinforcement learning algorithms
- average reward
- average cost
- risk sensitive
- partially observable
- reward function
- infinite horizon
- model based reinforcement learning
- action sets
- state and action spaces
- decision processes
- planning under uncertainty
- decision theoretic planning
- temporal reasoning
- factored mdps
- decision diagrams
- discounted reward
- temporal logic
- action space
- modal logic
- inference rules
- multi agent systems
- state abstraction
- sufficient conditions
- markov decision process
- search algorithm
- multi agent