ISAACS: Iterative Soft Adversarial Actor-Critic for Safety.
Kai-Chieh HsuDuy Phuong NguyenJaime Fernández FisacPublished in: CoRR (2022)
Keyphrases
- actor critic
- reinforcement learning
- policy gradient
- approximate dynamic programming
- optimal control
- temporal difference
- neuro fuzzy
- gradient method
- reinforcement learning algorithms
- multi agent
- machine learning
- function approximation
- decision making
- optimal solution
- dynamic programming
- neural network
- importance sampling
- policy iteration
- average reward