Fully Bayesian Recurrent Neural Networks for Safe Reinforcement Learning

Open in new window