1
$\begingroup$

If A is conditionally independent of B and C given D. And that B and C are not conditionally independent given D. Can I write P(A,B,C|D) as P(A|D) * P(B,C|D)?

Thanks!

$\endgroup$
  • 1
    $\begingroup$ Do you mean $A$ is conditionally independent of $B$ and conditionally independent of $C$, or conditionally independent of $B\cap C$? $\endgroup$ – André Nicolas Jun 12 '14 at 0:26
  • $\begingroup$ I meant A is conditionally independent of B and conditionally independent of C. Sorry! $\endgroup$ – Mat Jun 12 '14 at 1:41
1
$\begingroup$

Yes! Although there is ambiguity what you mean, which AndréNicolas points out in the above comments, in either case the answer to your question is yes! The conditional independence of $A$ with $B$ and $C$ given $D$ (whether, you mean conditional independence with $B$ and conditional independence with $C$, or just conditional independence of $B \cap C$) is defined such that: $$P(A, B, C \mid D) = P(A \mid D) P(B,C \mid D)$$

Since $B$ and $C$ are not conditionally independent given $D$, you should also note that $P(B,C\mid D) \neq P(B\mid D) P(C \mid D)$.

$\endgroup$

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.