Off-Policy Meta-Reinforcement Learning With Belief-Based Task Inference.
Takahisa ImagawaTakuya HiraokaYoshimasa TsuruokaPublished in: IEEE Access (2022)
Keyphrases
- reinforcement learning
- evidential reasoning
- function approximation
- inference process
- state space
- belief functions
- meta level
- machine learning
- bayesian networks
- multi agent
- policy search
- optimal policy
- data sets
- bayesian inference
- transition model
- reinforcement learning algorithms
- probabilistic inference
- belief networks
- graphical models
- inference engine
- optimal control
- temporal difference
- random fields
- bayesian model
- temporal difference learning
- autonomous learning
- truth maintenance systems
- learning tasks
- meta reasoning