Learning as MAP Inference in Discrete Graphical Models

Open in new window