Enhancing Deep Reinforcement Learning with Executable Specifications.
Raz YerushalmiPublished in: ICSE Companion (2023)
Keyphrases
- reinforcement learning
- formal specification
- attribute grammars
- control flow
- state space
- function approximation
- markov decision processes
- high level
- specification language
- reinforcement learning algorithms
- optimal policy
- machine learning
- model free
- delay insensitive
- deep learning
- active learning
- robotic control
- temporal difference
- action selection
- database
- data flow
- optimal control
- model checking
- multi agent
- finite state machines
- context free grammars
- learning problems
- hidden markov models
- learning process
- learning algorithm
- real time