Policy Learning for Time-Bounded Reachability in Continuous-Time Markov Decision Processes via Doubly-Stochastic Gradient Ascent.
Ezio BartocciLuca BortolussiTomás BrázdilDimitrios MiliosGuido SanguinettiPublished in: QEST (2016)
Keyphrases
- markov decision processes
- state space
- optimal policy
- reinforcement learning
- partially observable
- gradient ascent
- infinite horizon
- policy iteration
- finite state
- stochastic games
- dynamic programming
- average reward
- policy gradient
- learning algorithm
- partially observable markov decision processes
- markov decision process
- state action
- reinforcement learning algorithms
- action space
- average cost
- decision problems
- markov chain
- supervised learning
- image segmentation