Domain Adaptation for Reinforcement Learning on the Atari.
Thomas CarrMaria ChliGeorge VogiatzisPublished in: AAMAS (2019)
Keyphrases
- domain adaptation
- reinforcement learning
- transfer learning
- cross domain
- function approximation
- target domain
- labeled data
- multiple sources
- semi supervised
- temporal difference
- sentiment classification
- semi supervised learning
- reinforcement learning algorithms
- learning tasks
- machine learning
- supervised learning
- document classification
- state space
- model free
- knowledge transfer
- test data
- optimal control
- markov decision processes
- active learning
- multi agent
- decision trees
- data sets
- action space
- control problems
- co training
- learning process
- neural network