Polynomial Time Algorithms for Branching Markov Decision Processes and Probabilistic Min(Max) Polynomial Bellman Equations.
Kousha EtessamiAlistair StewartMihalis YannakakisPublished in: ICALP (1) (2012)
Keyphrases
- markov decision processes
- min max
- policy iteration
- factored mdps
- computational complexity
- state space
- reinforcement learning
- reachability analysis
- dynamic programming
- optimal policy
- learning algorithm
- planning under uncertainty
- bayesian networks
- transition matrices
- probabilistic planning
- infinite horizon
- finite state
- probabilistic model
- markov decision process
- average cost
- reinforcement learning algorithms
- decision theoretic planning