28f0b864598a1291557bed248a998d4e-Reviews.html

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. This paper extends the work of Yang et al (2012) on undirected graphical models where the node-wise conditional distributions are from the univariate exponential family; in this paper, the authors allow covariates to be included and also allow the distribution type to be node specific. They derive the form of the natural parameter for the response distributions under the assumption that the node-wise conditional distributions of the covariates are also from the univariate exponential family, and prove guarantees on the learnability of the models. Finally, the authors present some limited experimental results on synthetic and real data. The paper is well written, particularly given the density of material.