Gradient Shaping for Multi-Constraint Safe Reinforcement Learning.
Yihang YaoZuxin LiuZhepeng CenPeide HuangTingnan ZhangWenhao YuDing ZhaoPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- reward shaping
- reinforcement learning algorithms
- model free
- markov decision processes
- learning algorithm
- policy gradient
- optimal policy
- state space
- dynamic programming
- machine learning
- function approximation
- action selection
- constraint networks
- image gradient
- complex domains
- multi agent reinforcement learning