Avoiding fusion plasma tearing instability with deep reinforcement learning.
Jaemin SeoSangKyeun KimAzarakhsh JalalvandRory ConlinAndrew RothsteinJoseph AbbateKeith EricksonJosiah WaiRicardo ShoushaEgemen KolemenPublished in: Nat. (2024)
Keyphrases
- reinforcement learning
- function approximation
- state space
- information fusion
- temporal difference
- data fusion
- high energy
- fusion method
- thin film
- multi sensor
- reinforcement learning algorithms
- image registration
- dynamic programming
- learning process
- fusion algorithm
- temporal difference learning
- model free
- sensor fusion
- fusion framework
- optimal control
- image fusion
- optimal policy
- robotic control
- fusion process
- electric field
- learning algorithm
- fusion methods
- action selection
- high density
- multi agent
- machine learning
- deep learning
- function approximators
- learning agents
- reward function
- reinforcement learning methods
- stochastic approximation
- autonomous learning
- multi agent reinforcement learning
- fusion model
- transition model
- policy search