0
$\begingroup$

The question says provided a random sample $(x_1, x_2, \ldots ,x_n)=x$ from a Poisson distribution say $P(\theta)$.

It asks to find the Jeffreys prior distribution for $\theta$ and then find the posterior distribution of $\theta|x$.

I found the Jeffreys prior but have a doubt on the 2nd part of the question. What I know is given a Poisson prior, we would find the posterior distribution which will usually be a Gamma distribution. But they didn't provide other distributions in the question. How can we find the posterior distribution?

$\endgroup$
1
$\begingroup$

The Jeffreys' (improper) prior for a $\operatorname{Poisson}(\theta)$ is $$ p_{\mathrm{prior}}(\theta)\propto \theta^{-1/2} 1_{\theta>0}. $$ You are given $x_i\sim\operatorname{Poisson}(\theta)$, so assuming $x_i$s are independent(!), the posterior is $$ p_{\mathrm{posterior}}(\theta\mid x)\propto p(x\mid\theta)p_{\mathrm{prior}}(\theta) \propto e^{-n\theta}\theta^{-1/2+\sum x_i}1_{\theta>0} $$ i.e., $\theta\mid x\sim\operatorname{Gamma}(\alpha=\frac12+\sum x_i,\beta=n)$.

$\endgroup$

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.