4
$\begingroup$

Let $X_1+X_2+X_3+X_4$ be iid exponential random variables with parameter λ, and $S=X_1+X_2+X_3+X_4$
S follows the gamma distribution with parameters $\lambda$ and $r=4$.
We know that an exponential random variable X with parameter $\lambda$ has moment generating function
$E[e^{tX}]=\frac{\lambda}{\lambda-t}$ if $t<\lambda$ and $+\infty$ if $t \geq \lambda$
a) find MGF $M_S(t)$ (don't forget to declare the domain)
b) find the 1st and 2nd moment of S, then find the variance
I know that for b, you can just find $M_S'(0)$ and $M_S''(0)$ but for a, I'm not sure on how to get started on finding the moment generating function

$\endgroup$
  • $\begingroup$ Look up the solution for the sum of $n$ exponentials (google is your friend) $\endgroup$ – wolfies Dec 10 '15 at 17:16
1
$\begingroup$

The mgf of a sum of independent random variables is the product of the individual mgf.

In our case each $X_i$ has mgf $\frac{\lambda}{\lambda-t}$, and therefore the sum has mgf $\left(\frac{\lambda}{\lambda-t}\right)^4$ for $t\lt \lambda$.

You know how to do b) using the mgf. As a check, note that the mean of $S$ is the sum of the means of the $X_i$, and the variance of $S$ is the sum of the variances of the $X_i$.

$\endgroup$

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.