This is a short shout out to Prof. Dr. Philipp Hennig, the course instructor for Tübingen Universities Probabilistic Machine Learning course available on YouTube over here.
To give you some background, we took Daphne Koeller’s Probabilistic Graphical Model course last year on Coursera, so we can compare Dr Phillip’s course with hers.
Without a doubt, Dr Phillip’s course is better. For example, I was unclear on how message parsing even arose in Graphical models and Dr Phillip explained it very well. He also gave a very good connection between Graphical Models and the statistical physics domain.
His presentation style is clear and methodical and is catered to a technically fluent audience.
But be forewarned, you will need to be comfortable with Mathematical notation, especially set notation. With that out of the way, you can enjoy his well thought out lecture.
This is called a Spalier in German.
Watch Lecture 17 for the connection.