Transfer RL across Observation Feature Spaces via Model-Based Regularization.
Yanchao SunRuijie ZhengXiyao WangAndrew E. CohenFurong HuangPublished in: ICLR (2022)
Keyphrases
- feature space
- model free
- reinforcement learning
- transfer learning
- high dimensional
- function approximation
- dimensionality reduction
- polynomial kernels
- training samples
- feature vectors
- multi agent
- markov decision processes
- reinforcement learning algorithms
- image representation
- knowledge transfer
- nearest neighbor search
- mercer kernels
- low dimensional
- training set
- learning algorithm
- labeled data
- prior information
- hyperplane
- state space
- dynamic programming
- cross domain
- active learning
- kernel matrix
- action selection
- feature selection
- learning agents
- machine learning