Show that convergence in the mean implies convergence of the means

730 Views Asked by At

Question:

Let $X_n$, n = 1,... denote a sequence of real-valued random variables; $X_n$ is said to converge in mean if

$\hspace{20mm}$$$\lim_{n\to\infty} E[|X_n-X|] = 0$$

Show that if $X_n$ converges to X in mean and $E[|X|] < \infty$, then

$\hspace{20mm}$ $$\lim_{n\to\infty} E[X_n] = E[X]$$

I'm not entirely sure about this one, but I think the triangle inequality can be used:

$\hspace{20mm}$$$\lim_{n\to\infty}E[|X_n-X|] \geq \lim_{n\to\infty}E[|X_n|]-E[|X|] = 0$$

1

There are 1 best solutions below

0
On

Use the following integral property :

$$\big\vert \mathrm{E}\left[ Y \right] \big\vert \leq \mathrm{E}\left[ \vert Y \vert \right]$$

By taking ($Y=X_{n}-X$) and using the linearity of the expectation, we get :

$$\big\vert \mathrm{E}\left[ X_{n} \right] - \mathrm{E}\left[ X \right] \big\vert = \big\vert \mathrm{E}\left[ X_{n}-X \right] \big\vert \leq \mathrm{E}\left[ \big\vert X_{n} - X \big\vert \right]$$

Since we know that $\displaystyle \lim \limits_{n \to +\infty} \mathrm{E}\left[ \big\vert X_{n}-X \big\vert \right] = 0$, we get :

$$ \lim \limits_{n \to +\infty} \mathrm{E}\left[X_{n}\right] = \mathrm{E}\left[X\right] $$