Reinforcement Learning of Control Policy for Linear Temporal Logic Specifications Using Limit-Deterministic Büchi Automata.
Ryohei OuraAmi SakakibaraToshimitsu UshioPublished in: CoRR (2020)
Keyphrases
- control policy
- linear temporal logic
- reinforcement learning
- bounded model checking
- deterministic automata
- finite state automaton
- model checking
- temporal logic
- transition systems
- approximate dynamic programming
- state space
- finite state
- turing machine
- control policies
- formal specification
- average cost
- long run
- function approximation
- optimal policy
- finite automata
- concurrent systems
- formal verification
- markov decision processes
- fully observable
- specification language
- reactive systems
- partially observable
- dynamic programming
- learning algorithm
- action selection
- orders of magnitude
- regular expressions
- markov decision problems
- model free
- control flow
- machine learning