Given the recent success of machine learning in solving computational tasks previously thought unsolvable, we propose to develop methods that apply Bayesian learning to perform meta-learning to speed ...