Meta Reinforcement Learning with Autonomous Inference of Subtask Dependencies.
Sungryull SohnHyunjae WooJongwook ChoiHonglak LeePublished in: ICLR (2020)
Keyphrases
- reinforcement learning
- autonomous learning
- function approximation
- bayesian networks
- relational dependency networks
- learning capabilities
- cooperative
- markov decision processes
- statistical dependencies
- probabilistic inference
- supervised learning
- robotic control
- learning algorithm
- dependency analysis
- temporal difference
- inference engine
- neural network
- meta reasoning
- policy search
- transition model
- multi agent reinforcement learning
- action selection
- meta level
- relational data
- domain specific
- probability distribution
- expert systems
- multi agent
- bayesian inference
- optimal control
- belief networks
- learning problems
- optimal policy
- graphical models