Beating the Likelihood: Marginalization-Based Parameter Learning in Graphical Models

Likelihood based-learning of graphical models faces challenges of computational-complexity and robustness to model error. This paper studies methods that fit parameters directly to maximize a measure of the accuracy of predicted marginals, taking into account both model and inference approximations at training time. Experiments on imaging problems suggest marginalization-based learning performs better than likelihood-based approximations on difficult problems where the model being fit is approximate in nature.