Conditional probability and its expectation of continuous random variables

525 Views Asked by At

I have two subtle questions on the definition of conditional probability and its expectation. Here's the thing.

Question 1.

Definition of a conditional probability density function

$f_{X|Y}(x,y)=\frac{f(x,y)}{f_{Y}(y)}$

And if I want to compute below probability:

$P(0<X<1|0<Y<2)$

then I think below equation is one to calculate:

$P(0<X<1|0<Y<2)=\frac{\int_{0}^{1}\int_{0}^{2}f(x,y)dydx}{\int_{0}^{2}f_{Y}(y)dy}$

and come to think of it. How about this:

$P(0<X<1|0<Y<2)=\int_{0}^{1}\int_{0}^{2}f_{X|Y}(x|y)dydx$

But it is easy to show that

$\frac{\int_{0}^{1}\int_{0}^{2}f(x,y)dydx}{\int_{0}^{2}f_{Y}(y)dy} \neq \int_{0}^{1}\int_{0}^{2}f_{X|Y}(x|y)dydx$

Where are the parts that I am missing at all?

Question 2.

The conditional expectation of $X$, given that $Y=y$, is defined for all values of $y$ such that $f_Y(y)>0$, by

$E[X|Y=y]=\int_{-\infty}^{\infty}xf_{X|Y}(x|y)dx$

Here's my question.

Since we are dealing with continuous random variable cases, we've learned that a value of a pdf at a particular point is meaningless since $P(X=a)=0$, where $X$ is a continuous one.

But, as you can see in the definition of conditional expectation, the given condition is $Y=y$ not that $Y$ is in some interval or something like that.

I don't see why it is a valid definition.

p.s. The text book is Introduction to Probability Models 10th Ed. by Sheldon M. Ross.