233509073ed3432027d48b1a83f5fbd2-Reviews.html

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. Reaction to the author feedback: I cannot agree with point (3): The work of Siracusa III and Fisher (AISTATS 2009) does *not* assume the data are iid, and it allows the generating stucture vary (even if only in a small set of graphical models). While the present proposed method is, in some sense, even more flexible, I'd find it misleading to claim that the proposed method is the first to address the problem. The proposed method is specifically targeted to scenarios where the data generating graphical model may change relatively frequently. The performance of the method is demonstrated using both simulated and real data.