I really should get around to reading that. His info theory book/lectures was/were very good. http://www.inference.phy.cam.ac.uk/itprnn/book.html