Learning resources for Bayesian Dynamic Networks? Increasingly, I've stumbled on the term Bayesian Dynamic Network(s). The field seems to be at the intersection of probabilistic graphical models, time series, Kalman filters, etc. Because there's so much depth to it, I'm having a hard time finding a resource that motivates these complex ideas from simpler concepts. Could anyone link a course, book, etc that builds these ideas "from the ground up"?
 A: The following lecture video, "L10 - Gaussian graphical models and Ising models", here of the course CMU Probabilistic Graphical Models Spring 2019 has a small section towards the end about modelling the structure of time varying dynamic Bayesian network; after covering how to infer the structure of a Gaussian graphical model using LASSO (which is a result that appeared in the Annals of Stats). You can find course materials e.g. problem sets, lecture slides etc. at the course page here - that year's iteration has a complete set of problem sheets and programming questions.
For graphical models, approximate inference methods such as MCMC/variational inference, and the applicability of graphical models to deep learning (e.g. deep graphical models), I found this course has given me literacy on the more exotic techniques encountered in papers.
For self-study, I highly recommend - I have been working through it and has been very enjoyable, especially the problem sets. As a point of preference, or study disposition, I don't think I would have been able to cultivate the same knowledge if I had just read a textbook such as Daphne Koller's - the textbook is incredibly dense, and whilst comprehensive and rigorous, is not clear with the big picture. There is also arguably a much clearer unpublished textbook on graphical models written by Michael Jordan, bits of which the course uses. There are invaluable points the instructor covers which you cannot find in textbooks, and they are able to communicate empahsis and give a big picture view that means you don't get lost agonising over minor details (like I do when I read heavier maths books).
