Discounted Continuous-Time Markov Decision Processes with Unbounded Rates: The Convex Analytic Approach.
Alexey B. PiunovskiyYi ZhangPublished in: SIAM J. Control. Optim. (2011)
Keyphrases
- markov decision processes
- state space
- stationary policies
- optimal policy
- finite state
- infinite horizon
- average reward
- dynamic programming
- average cost
- finite horizon
- reinforcement learning
- optimal control
- markov chain
- decision theoretic planning
- policy iteration
- markov decision process
- transition matrices
- reachability analysis
- model based reinforcement learning
- action space
- planning under uncertainty
- partially observable
- dynamical systems
- reinforcement learning algorithms
- factored mdps
- decision processes
- semi markov decision processes
- decision making
- risk sensitive
- discounted reward
- stochastic processes
- interval estimation
- state abstraction
- inventory level
- action sets
- reward function
- heuristic search
- data mining