The words contained in this file might help you see if this file matches what you are looking for:

...Eea statsa information theory lecture method of types lecturer tsachy weissman scribe fang cai rob jones yi sun can wang in last we framed the problem lossy compression and gave main theorem that characterizes tradeo between rate distortion context with as our motivation this week are going to talk about expand tools related typical sequences notion strong typicality conditional which not only interesting their own right but also serve very well when go back establish result some concrete schemes for how they clustering machine learning problems today notation denote xn x xi r n a i p empirical distribution type class denition is probability vector denotes collection all length e or t example if then pn b c following show number dierent classes induced by be upper bounded something polynomial doesn increase exponentially proof everyempiricaldistributionp determined wheren means times symbol appears sequence since each take on no more than values thus have element therefore there at mos...