I am working on this question:

Prove that $X_{n}\rightarrow 0$ a.s. if and only if for every $\epsilon>0$, there exists $n$ such that the following holds: for every random variable $N:\Omega\rightarrow\{n,n+1,\cdots\}$, we have $$P\Big(\{\omega:|X_{N(\omega)}(\omega)|>\epsilon\}\Big)<\epsilon.$$

Is this question equivalent to asking me to prove "almost surely convergence to $0$ if and only if convergence to $0$ almost surely"?

If so, the direction $(\Rightarrow)$ can be proved following this: Convergence in measure and almost everywhere

However, isn't the direction $(\Leftarrow)$ not generally true? I can surely prove that there exists a subsequence $X_{k_{n}}$ of $X_{n}$ converges to $0$ almost surely...

Could someone tell me what this question is really asking about? I don't really want to spend time proving a wrong thing..

Thank you!


First, $X_n\to 0$ a.s. iff for any $\epsilon>0$, there exists $n\ge 1$ s.t. $\mathsf{P}(\sup_{m\ge n}|X_m|>\epsilon)<\epsilon$. In the following we fix $\epsilon>0$.

$(\Rightarrow)$ Suppose that $X_n\to 0$ a.s. Then since for any r.v. $N$ on $\{n,n+1,\ldots\}$, $|X_{N}|\le \sup_{m\ge n}|X_m|$, the result follows from the above statement.

$(\Leftarrow)$ Let $N_n':=\inf\{m\ge n:|X_m|>\epsilon\}$. Define $N_n:=N_n'1\{N_n'<\infty\}+n1\{N_n'=\infty\}$. Then $\{\sup_{m\ge n}|X_m|>\epsilon\}=\{|X_{N_n}|>\epsilon\}$. However, there exists $n\ge 1$ s.t. $\mathsf{P}(|X_{N_n}|>\epsilon)<\epsilon$.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ – Aloizio Macedo Sep 26 '19 at 22:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.