1
$\begingroup$

enter image description hereI am confused in obtaining a term in Bayesian Formula. I have attached my partial solution in the image. How can I calculate the term P(X1/X3)??. Image attached

$\endgroup$
  • $\begingroup$ I think it should be $0.5$. $\endgroup$ – Ryan Aug 7 '17 at 6:11
  • $\begingroup$ Could you please explain it. Thank you @Ryan $\endgroup$ – Abdul Karim Khan Aug 7 '17 at 6:13
  • $\begingroup$ I have already attached answer for you, please check :-) $\endgroup$ – Ryan Aug 7 '17 at 6:18
0
$\begingroup$

Hint: The most important part is to solve $p(A|x_1)$ and $p(\neg A|x_1)$.

$$p(A|x_1) = \frac{p(x_1|A)p(A)}{p(x_1)} = \frac{p(x_1|A)p(A)}{p(x_1|A)p(A) + p(x_1|\neg A)p(\neg A)} = 0.25$$

Then, $p(\neg A|x_1) = 0.75$.

(PS: Based on the Chapman-Kolmogorov equation, the result of $p(x_3|x_1)$ is obvious from $p(A|x_1)$ and $p(\neg A|x_1)$. Here, we still explain it in details.)

Note that $p(x_3|x_1) = p(x_3, A|x_1) + p(x_3, \neg A|x_1)$ (by the law of total probability). We calculate the following two terms

$$p(x_3,A|x_1) = p(x_3|A)p(A|x_1) = 0.05,$$ $$p(x_3,\neg A|x_1) = p(x_3|A)p(\neg A|x_1) = 0.45,$$

where $p(x_3|A,x_1) = p(x_3|A)$ and $p(x_3|\neg A,x_1) = p(x_3|\neg A)$.

Now, $p(x_3|x_1) = p(x_3, A|x_1) + p(x_3, \neg A|x_1) = 0.05 + 0.45 = 0.5$.

$\endgroup$
  • $\begingroup$ Would you please explain why p(x3|x1)=p(x3,A|x1)+p(x3,¬A|x1)? @Ryan $\endgroup$ – Abdul Karim Khan Aug 7 '17 at 6:27
  • $\begingroup$ Sorry, It would not be that complex. Just the law of total probability en.wikipedia.org/wiki/Law_of_total_probability, which is the Chapman-Kolmogorov equation based on. $\endgroup$ – Ryan Aug 7 '17 at 6:35
  • $\begingroup$ Yeah, Thanks. I solved it using law of total probability.@Ryan $\endgroup$ – Abdul Karim Khan Aug 7 '17 at 6:37
  • $\begingroup$ @AbdulKarimKhan Cheers! $\endgroup$ – Ryan Aug 7 '17 at 6:38

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.