8
$\begingroup$

I only saw Bayes rule with one conditioned event and now I need to work with two events and I am not sure if this is a right way to do it:

$$ p(A \mid B, C) = \frac{p(B \mid A, C)\ p(A \mid C)}{p(B \mid C)} = \frac{p(B, A \mid C)}{p(B\mid C)}$$

I honestly do not know if this makes sense, but I needed to have $p(B\mid C)$ in denominator for the problem that I am solving, so I thought this might be the way to do it.

Why is the above true or false?

$\endgroup$

1 Answer 1

8
$\begingroup$

This is valid. The key is that $B, C = B \cap C$ is a single event (that is, a subset of the sample space), so your result follows from the definition of conditional probability. (Note that we'll use the same trick with the event $A, B = A \cap B$.) $$p(A \mid B, C) = \frac{p(A, B, C)}{p(B, C)} = \frac{p(A, B \mid C) p(C)}{p(B \mid C) p(C)}$$ This approach can also be used to show that your middle term, $$\frac{p(B \mid A, C) p(A \mid C)}{p(B \mid C)},$$ is also equivalent to each of these expressions.

$\endgroup$

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.