The Provable Benefit of Unsupervised Data Sharing for Offline Reinforcement Learning.
Hao HuYiqin YangQianchuan ZhaoChongjie ZhangPublished in: ICLR (2023)
Keyphrases
- data sharing
- reinforcement learning
- data access
- data integration
- distributed data
- peer to peer
- supervised learning
- information sharing
- metadata management
- private data
- semantic interoperability
- state space
- security administration
- government agencies
- data privacy
- machine learning
- learning algorithm
- sensitive data
- ontology mapping
- data mining
- databases
- data sets