Tag: A Bayesian Approach to Robust Reinforcement Learning