I want to prove that $X,Y$ scalar random variables are independent iff conditional regular distribution of $X|Y$ is almost surely the same as the distribution of $X$. A simple result to prove for the... "regular" conditional distribution.

The definition employed for the regular distribution is the following. Let $(\Omega,\mathcal F,P)$ be a probability space, and $(\Omega',\mathcal F')$ a measurable space. Now $\mu:\Omega\times\mathcal F'\rightarrow \mathbb R$ is the regular conditional distribution if:

1) For every $A\in \mathcal F'$, $\mu(.,A)=E(I_{X\in A}|\mathcal G)$ almost surely. $\mathcal G\subseteq \mathcal F$, and in this case $\mathcal G=\sigma (Y)$, which is the pre-image of the borel sigma algebra.

2) For almost every $\omega\in \Omega$, $\mu(\omega,.)$ is a probability measure.

This definition I find very hard to use in order to prove almost anything.


Take $A=(-\infty , x]$. If $X$ and $Y$ are independent then $\mu(\omega, A)=E(I_{(X \in A)} |Y)=P(X^{-1}(A))=P\{X\leq x\}$ so the conditional distribution coincides with teh distribution of $X$. Conversely if this condition holds integrate both sides over $\{Y\leq y\}$ to get $P\{X\leq x,Y\leq y\} =P\{X\leq x\}P\{Y\leq y\}$.

| cite | improve this answer | |

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.