Safe Imitation Learning via Fast Bayesian Reward Inference from Preferences.

Daniel S. Brown Russell Coleman Ravi Srinivasan Scott Niekum

Published in: CoRR (2020)

Keyphrases

imitation learning
reinforcement learning
bayesian inference
bayesian networks
maximum margin
humanoid robot

robotic systems
probabilistic inference
relational domains
maximum likelihood
relational databases
posterior probability

belief networks
learning algorithm
reinforcement learning methods
probabilistic reasoning
posterior distribution
support vector

dynamic bayesian networks
multi relational
function approximation
relational data
multi modal
feature space