Optimal Control of Discounted-Reward Markov Decision Processes Under Linear Temporal Logic Specifications.
Krishna Chaitanya KalagarlaRahul JainPierluigi NuzzoPublished in: ACC (2021)
Keyphrases
- optimal control
- discounted reward
- linear temporal logic
- markov decision processes
- dynamic programming
- policy iteration
- reinforcement learning
- average reward
- temporal logic
- model checking
- infinite horizon
- control problems
- optimal policy
- risk sensitive
- finite state
- average cost
- state space
- control strategy
- reinforcement learning algorithms
- planning under uncertainty
- markov decision process
- partially observable
- decision theoretic planning
- reward function
- state and action spaces
- function approximation
- hierarchical reinforcement learning
- machine learning
- markov decision problems
- decision processes
- action space
- classical planning
- multistage