On the (In)Tractability of Reinforcement Learning for LTL Objectives.
Cambridge YangMichael L. LittmanMichael CarbinPublished in: IJCAI (2022)
Keyphrases
- reinforcement learning
- function approximation
- temporal logic
- model checking
- linear temporal logic
- constraint satisfaction
- bounded model checking
- temporal difference learning
- reinforcement learning algorithms
- state space
- optimal policy
- markov decision processes
- np complete
- computational complexity
- optimal control
- machine learning
- supervised learning
- dynamic programming
- robotic control
- learning process
- multi agent reinforcement learning
- deterministic automata
- transition model
- reinforcement learning methods
- neural network
- reward function
- temporal difference
- model free
- learning problems
- heuristic search
- np hard
- multi agent