Learning Action Translator for Meta Reinforcement Learning on Sparse-Reward Tasks.
Yijie GuoQiucheng WuHonglak LeePublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- action selection
- learning process
- learning algorithm
- state action
- function approximation
- supervised learning
- learning agent
- state space
- optimal control
- partially observable environments
- reward shaping
- autonomous learning
- reinforcement learning methods
- reinforcement learning algorithms
- machine learning
- multi agent
- learning agents
- partially observable domains
- high dimensional
- dynamic programming
- policy gradient
- mobile robot
- action models
- temporal difference learning
- learning problems
- online learning
- function approximators
- optimal policy
- real robot
- learning capabilities
- transfer learning
- markov decision processes