I'm studying bayesian stats on my own and came across the following problem on Coursera. Can anyone help me understand how to work through this?

$x_{i}\stackrel {iid}{\sim} Beta(\alpha,\beta), i=1....n $

$ \alpha \sim Gamma(a,b)$

$ \beta \sim Gamma(r,s)$

Where $\alpha$ and $\beta$ are independent a priori. Give the full conditional density for $\alpha$ up to proportionality $p(\alpha|\beta,x)$?

Edited: I'm getting stuck with the setup for this.

I have assumed that $p(\alpha|\beta,x)$? = $\frac{p(\alpha\cap\beta\cap x)}{p(\beta\cap x)}$

I'm not sure how to proceed from here. Any help will be highly appreciated.


  • $\begingroup$ Thanks Xi'an. I'm probably going down the wrong path with this. I have edited the question to explain where I'm possibly getting confused. $\endgroup$ – Electromagnet Jan 4 at 7:26
  • 1
    $\begingroup$ The notation $p(\alpha\cap\beta\cap x)$ is only appropriate for events and sets, not for random variables:$$p(\alpha|\beta,x)=p(\alpha,\beta,x)\big/p(\beta,x)$$ $\endgroup$ – Xi'an Jan 4 at 7:32
  • $\begingroup$ Write down$$p(\alpha,\beta,x)=p(\alpha)p(\beta)\prod_i p(x_i|\alpha,\beta)$$and turn it into a density on $\alpha$. $\endgroup$ – Xi'an Jan 4 at 7:42

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.