2

Could you help me prove that the following sequence of independent random variables satisfy Weak Law of Large Numbers?

\begin{equation} P(x_k=\pm 2^k)=\frac{1}{2^{2k+1}} \end{equation}

\begin{equation} P(x_k=0)=1-\frac{1}{2^{2k}} \end{equation}

I made this,

\begin{equation} E(x_k)=(2^k)(\frac{1}{2^{2k+1}})+(-2^k)(\frac{1}{2^{2k+1}})+(0)\left( 1-\frac{1}{2^{2k}}\right)=0< \infty \end{equation} also,

\begin{equation} P(|\frac{S_n}{n}-0|> \epsilon)=P(|\frac{S_n}{n}|> \epsilon)\leq \frac{V(S_n/n)}{\epsilon^2} \end{equation} also,

\begin{equation} V(\frac{S_n}{n})=\frac{E(x_k^2)}{n} \end{equation}

and then I don't know what to do.

DB1
  • 21
  • The next step is to compute $E(x_k^2).$ That's super easy because there are only two possible values of $x_k^2.$ For additional guidance, emulate the answers at https://stats.stackexchange.com/questions/187285/chebychev-s-weak-law-of-large-numbers, of which this problem is a special case. – whuber Aug 07 '20 at 18:22

0 Answers0