I gave two lectures on computational mechanics, on 18 and 19 June 1998, to students from the complex systems summer school. The first was mostly about causal states, and proving various optimality results about them, and I was quite proud of it. The second was about estimating causal states from actual data, and wasn't nearly so good. I may get around to texing up my notes for the second lecture, but for now what's on-line is the first lecture and the reading list.
---CRS