1
$\begingroup$

I was doing this exercise and then I checked the solution, but I got the solution wrong.

$X_{1},...X_{n} \sim^{iid} N(0,\theta)$, i.e.

$f_{X_{i}} (x)= \frac{1}{\sqrt{2 \pi \theta}} -e^\frac{x^2}{2 \theta}$

b) find the CRLB for the class of unbiased estimators of $ \theta$

So I compute the Likelihood and the logarithm, then I take first and second derivative and the calculation then should be:

$ I =-E(\frac{1}{2\theta^2} - \frac{x^2}{\theta^3}$) and then CRLB = $\frac{1}{nI}$

However, doing this I came to the wrong result, can help me understand why?

$\endgroup$
  • $\begingroup$ Precisely what result did you obtain? $\endgroup$ – whuber Sep 5 '18 at 11:52
1
$\begingroup$

There is no problem in your derivation, however you can have a closed-form result, which isn't in your answer, and probably that is why you got your points deducted.

You can further develop

$$ I(\theta)=-\mathbb{E}(l^{''}\left(X\vert\theta\right))=-\mathbb{E}\left(\frac{1}{2\theta^2}-\frac{X^2}{\theta^3}\right)=\frac{1}{2\theta^2} $$

and therefore the CRLB is $$ CRLB=\frac{2\theta^2}{n} $$

$\endgroup$

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.