I’m at the 2011 School of Information Theory at UT Austin — I’m mostly here to help out with the activities on Monday, which I will blog about after they happen (it’s a secret!) I missed most of the talks, but I went to the first part of Bob Gray’s lecture this morning on stationary codes and the connections between information theory and ergodic theory. Here’s his set of inclusions of the different species of random processes:

IID B-processes K-processes strongly mixing weakly mixing stationary ergodic stationary block stationary asymptotically stationary asymptotically mean stationary

B-processes are the result of applying stationary codes to IID process and K-processes are those which satisfy the Kolmogorov 0-1 law. The last one implies that sample averages converge. It’s only appropriate to mention it, given the title of this blog.

The School is a nice event with a mixture of classical and “hot topics” in information theory. Even though my own research interests have shifted a bit away from the hard core of IT, there are still fun things to think about and it’s nice to see some new new and new old results.