Deep reinforcement learning uncovers processes for separating azeotropic mixtures without prior knowledge.
Quirin GöttlJonathan PirnayJakob BurgerDominik G. GrimmPublished in: CoRR (2023)
Keyphrases
- prior knowledge
- reinforcement learning
- reinforcement learning methods
- multi agent
- learning process
- function approximation
- markov decision processes
- reinforcement learning algorithms
- computational models
- prior information
- dynamic programming
- model free
- labeled data
- neural network
- maximum likelihood
- latent variables
- state space
- shape prior
- optimal control
- temporal difference
- domain knowledge
- training data