Predictable Reinforcement Learning Dynamics through Entropy Rate Minimization.
Daniel Jarne OrniaGiannis DelimpaltadakisJens KoberJavier Alonso-MoraPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- objective function
- function approximation
- dynamic model
- mutual information
- information entropy
- reinforcement learning algorithms
- model free
- state space
- machine learning
- optimal control
- information theory
- neural network
- markov decision processes
- decision trees
- action selection
- learning algorithm
- robot control
- action space
- multi agent reinforcement learning