Robust Task Representations for Offline Meta-Reinforcement Learning via Contrastive Learning.
Haoqi YuanZongqing LuPublished in: ICML (2022)
Keyphrases
- semi supervised
- supervised learning
- reinforcement learning
- active learning
- learning algorithm
- learning process
- learning problems
- online learning
- multi agent
- learning tasks
- machine learning
- learning systems
- evolutionary learning
- autonomous learning
- learning capabilities
- policy search
- function approximators
- real time
- reinforcement learning algorithms
- knowledge acquisition
- object recognition
- neural network
- data sets