Universal Successor Representations for Transfer Reinforcement Learning.
Chen MaJunfeng WenYoshua BengioPublished in: ICLR (Workshop) (2018)
Keyphrases
- reinforcement learning
- transfer learning
- function approximation
- state space
- learning algorithm
- sequence prediction
- multi agent
- machine learning
- learning process
- optimal policy
- higher level
- distributed representations
- multi agent reinforcement learning
- multiple representations
- function approximators
- action selection
- learning tasks
- neural network
- symbolic representation
- active learning
- artificial neural networks
- decision trees
- previously learned
- website
- information systems
- real world