2
$\begingroup$

If parameter p ~ Beta(1,1), this would reflect we know nothing about parameter $p$. Generalizing to the multivariate case, how would the same be said about a vector $P$ of probability parameters $p_i$?

$\endgroup$
3
  • 1
    $\begingroup$ "this would reflect we know nothing about parameter $p$" — it is a uniform prior, but it's not an uninformative prior. $\endgroup$
    – Neil G
    Commented Dec 5, 2013 at 8:42
  • $\begingroup$ but would this not mean $p$ could be anything with equal probability? $\endgroup$ Commented Dec 5, 2013 at 19:31
  • $\begingroup$ Yes, but what if you had used odds $o \in [0, \infty)$ instead of probability $p$. Would you not want to choose a uniform prior for that? $\endgroup$
    – Neil G
    Commented Dec 5, 2013 at 21:52

1 Answer 1

11
$\begingroup$

A Beta distribution is just a special case of the Dirichlet distribution, that is, a Beta distribution is a Dirichlet distribution with two parameters, alpha and beta.

Dirichlet is the multidimensional generalisation (of Beta) with 'n' parameters instead of two. The parameters of Dirichlet are denoted by alpha with an index as a subscript. Setting all the alphas of a Dirichlet to 1 (no matter how many dimensions we are taking about) will give us the 'n' equivalents of Beta(1,1).

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.