Safe Reinforcement Learning Using Advantage-Based Intervention.
Nolan WagenerByron BootsChing-An ChengPublished in: ICML (2021)
Keyphrases
- reinforcement learning
- function approximation
- temporal difference learning
- reinforcement learning algorithms
- direct policy search
- learning algorithm
- computer vision
- control problems
- neural network
- temporal difference
- model free
- markov decision processes
- state space
- database systems
- least squares
- optimal control
- objective function
- image sequences
- decision trees
- function approximators
- learning agents
- machine learning
- data sets