Talk:Typical set

(Weakly) typical sequences (weak typicality, entropy typicality)

What is $p$ here? If it is a probability measure, why is there a $\Pr$ later also used? How are the two related? — Preceding unsigned comment added by 71.194.136.165 (talk) 12:01, 24 September 2013 (UTC)

Explicit quantifiers

Reading this as someone who isn't a specialist but who is looking to learn something.

I'd love to see the quantifiers spelled out more explicitly in the examples section. It talks about "the sequence" of all ones $(1,1,...,1)$. That presumably fixes $n$ to be the length of the sequence. But then it says this sequence is not typical because of the properties of its logarithmic probability as $n$ goes to infinity. So $n$ is not fixed. So then I don't know what "the sequence" means. Maybe it's talking about the sequence of sequences $(1), (1,1), (1,1,1), \ldots$ but I don't think the definition makes any reference to sequences of sequences in this way. In fact, the definition is pretty confusing too as it talks about "the typical set" without mentioning the dependence on $\epsilon$. Surely it should be called an $\epsilon$-typical set or something like that. Why the "the"? Surely every single sequence $(1), (1,1), \ldots$ is in some $\epsilon$-typical set for some $\epsilon$.