Reinforced Imitation Learning by Free Energy Principle.
Ryoya OgishimaIzumi KarinoYasuo KuniyoshiPublished in: CoRR (2021)
Keyphrases
- free energy
- imitation learning
- belief propagation
- fixed point
- upper bound
- humanoid robot
- competitive learning
- robotic systems
- reinforcement learning
- maximum margin
- posterior distribution
- approximate inference
- graphical models
- sufficient conditions
- latent variables
- probabilistic inference
- neural network
- dynamical systems
- markov random field
- feature selection