Risk Sensitive Dead-end Identification in Safety-Critical Offline Reinforcement Learning.
Taylor W. KillianSonali ParbhooMarzyeh GhassemiPublished in: Trans. Mach. Learn. Res. (2023)
Keyphrases
- risk sensitive
- safety critical
- reinforcement learning
- model free
- optimal control
- markov decision processes
- dead end
- formal methods
- control policies
- fault tolerant
- reinforcement learning algorithms
- markov decision problems
- embedded systems
- utility function
- optimal policy
- function approximation
- agent architecture
- state space
- policy iteration
- search algorithm
- dynamic programming
- real time
- adaptive systems
- support systems
- optimality criterion
- temporal difference
- markov decision process
- control strategies
- decision making
- learning algorithm
- path planning algorithm
- artificial intelligence
- action space
- multi agent
- path planning
- machine learning
- expert systems