1
$\begingroup$

I am practicing probability questions for my exam. I came across this question.

enter image description here

joint pdf

X and Y are two random variables with joint PDF uniform over the shaded area in the image.

a) What would be joint PDF expression?

b) Marginal PDFs of X and Y

c) E[X + Y]

d) Let Z be the event {Y>1}. Find the conditional PDF fX|Z(x|Z)

e) Comparing the conditional PDF fX|Z(x|Z) with the marginal pdf of fZ(z), is it implied that the random variables X and Y are independent? Explain.

For part (a), do I need to find the equation of the triangles?

For part (b), what I understand is that once I have the expression for the joint pdf I have to integrate for all Y to get marginal of X and vice versa.

I am clueless about parts c, d and e.

$\endgroup$
  • $\begingroup$ I am unable to post the image because my reputation is less than 10. The link however works. $\endgroup$ – The.Shark Oct 25 '17 at 7:00
0
$\begingroup$

You're on the right track with a and b. For c, you can use linearity: $E[X+Y] = E[X]+E[Y]$. When you have the marginals these expectations are just single integrals. Alternatively, you can calculate $E[X+Y]$ from the joint distribution as

$$E[X+Y] = \int\int (x+y) f_{X,Y} (x,y) dx dy$$

For d, calculate $P(Z)$. Then the joint distribution of $X$ and $Y$ is given by the pdf $\frac{I_Zf_{X,Y}}{P(Z)}$, where $I_Z$ is the indicator for $Z$ (one where $Z$ is true, zero elsewhere). Calculate the marginal of $X$ from this.

In e, I think $f_X(x) ($marginal pdf of $X$) is meant instead of $f_Z(z)$ ($Z$ is an event, not a random variable). Notice that the distributions $f_{X|Z}$ and $f_{X}$, even if they're the same, don't imply that $X$ and $Y$ are independent. For independence you need $f_{X|Y} = f_X$ or equivalently $f_{X,Y} = f_Xf_Y$.

It is clear from the joint distribution that in fact $X$ and $Y$ aren't independent: if $Y=1$, $X$ must be $1$, but if $Y=2$, $X$ is uniform over $[0,1]$. Hence knowing $Y$ gives information about $X$.

$\endgroup$

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.