Proving the Asymptotic Equipartition Property (AEP) for i.i.d. random variables

67 Views Asked by At

I'm currently working on some assignment and I'm not sure if my solution would be considered as correct in any way:

Let $(X)_{n \gt 1}$ be a stochastic process of i.i.d. random variables with values in a finite set $\mathcal{X}$ and distribution $p(x)$. Let $\mu = \mathbb{E} (X_1)$. Given $\varepsilon \gt 0$, show that

1.) $\mathbb{P}\left(\left|-\frac{1}{n} \log_2 p(x_1, \ldots, x_n) - H(X_1)\right| \leq \varepsilon\right) \rightarrow 1$

2.) $\mathbb{P}\left( \left| \frac{1}{n} \sum^{n}_{i=1} x_i - \mu \right| \leq \varepsilon \right) \rightarrow 1$

for $n \rightarrow \infty$.

If I rewrite 1. as

$\mathbb{P}\left(\left|-\frac{1}{n} \sum^{n}_{i = 1} \log_2(p(x_i)) - \mathbb{E}(-\log_2p(x_1))\right| \leq \varepsilon \right) \rightarrow1$

both 1. and 2. seem to be true due to the Weak Law of Large Numbers.

I have the feeling though that this is a little bit too easy and there must be something more to it.