Deep Reinforcement Learning based control algorithms: Training and validation using the ROS Framework in CARLA Simulator for Self-Driving applications.
Óscar Pérez-GillRafael BareaElena López GuillénLuis Miguel BergasaCarlos Gómez HuélamoRodrigo GutiérrezAlejandro DíazPublished in: IV (2021)
Keyphrases
- reinforcement learning
- main contribution
- learning algorithm
- computational complexity
- theoretical analysis
- optimization problems
- state space
- worst case
- model free
- optimal policy
- machine learning algorithms
- supervised learning
- semi supervised
- probabilistic model
- control system
- active learning
- training set
- multi agent
- decision trees