Domain Adaptation for Reinforcement Learning on the Atari.
Thomas CarrMaria ChliGeorge VogiatzisPublished in: CoRR (2018)
Keyphrases
- domain adaptation
- reinforcement learning
- transfer learning
- cross domain
- function approximation
- target domain
- labeled data
- multiple sources
- state space
- semi supervised learning
- sentiment classification
- semi supervised
- test data
- control problems
- machine learning
- training data
- learning algorithm
- unlabeled data
- reinforcement learning algorithms
- learning tasks
- model free
- temporal difference
- machine learning algorithms
- dynamical systems
- optimal control
- document classification
- unsupervised learning
- text classification
- dynamic programming
- multi agent
- action selection