4
$\begingroup$

Let $Y_1$ and $Y_2$ be independent r.v.s, and $X$ be another random variable.

What is $E[X|Y_1, Y_2]$? Is $E[X|Y_1, Y_2]$ equivalent to $E[X|Y_1] \cdot E[X|Y_2]$?

More specifically, is there a formula for the density of $f_{X|Y_1,Y_2}$ just like $f_{X|Y_1}$$=$$f_{X,Y_1}/f_{Y_1}$ (assuming that regular conditions such as $f_{Y_1}>0$ are satisfied)?

$\endgroup$

1 Answer 1

7
$\begingroup$

What is $E[X|Y_1,Y_2]$?

Literally, the expectation of $X$ given the values of both $Y_1$ and $Y_2$.

"The average height (X) of a 12 year old (Y1) male (Y2)" is an example.

Is $E[X|Y_1,Y_2]$ equivalent to $E[X|Y_1]⋅E[X|Y_2]$?

No. Consider the trivial case where $X$ is independent of both $Y_1$ and $Y_2$. You'd be asking "Is $E[X]$ equivalent to $E[X]⋅E[X]\,$?"

Or in the "average height of a 12 year old male" case, it's asking "is that the same as the average height of a 12 year old times the average height of a male?" ... which is plainly not the case.

is there a formula for the density of fX|Y1,Y2 just like fX|Y1=fX,Y1/fY1 (assuming that regular conditions such as fY1>0 are satisfied)

Of course!

$f_{X|Y_1,Y_2}=f_{X,Y_1,Y_2}/f_{Y_1,Y_2}$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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