Light-weight probing of unsupervised representations for Reinforcement Learning.
Wancong ZhangAnthony GX-ChenVlad SobalYann LeCunNicolas CarionPublished in: CoRR (2022)
Keyphrases
- lightweight
- reinforcement learning
- supervised learning
- unsupervised learning
- function approximation
- multi agent
- model free
- learning algorithm
- temporal difference
- function approximators
- machine learning
- data driven
- higher level
- semi supervised
- dynamic programming
- optimal control
- symbolic representation
- learning process
- unsupervised manner
- multiple representations
- unsupervised methods
- end to end
- markov decision processes
- optimal policy