SURF: Semi-supervised Reward Learning with Data Augmentation for Feedback-efficient Preference-based Reinforcement Learning.
Jongjin ParkYounggyo SeoJinwoo ShinHonglak LeePieter AbbeelKimin LeePublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- semi supervised
- database
- data sets
- learning algorithm
- training data
- supervised learning
- learning process
- prior knowledge
- data sources
- background knowledge
- data collection
- data analysis
- data mining techniques
- labeled data
- active learning
- multi agent
- learning tasks
- machine learning
- reward function
- partially labeled
- multi view
- state space
- data points