Laplace Maximum Margin Markov Networks

Learning sparse Markov networks based on the maximum margin principle remains an open problem in structured prediction. In this paper, we proposed the Laplace max-margin Markov network (LapM3N), and a