1
$\begingroup$

Let $X_1$, $X_2$, . . . , $X_n$ be a random sample from a $Uniform(θ, 1)$ population, where $θ < 1$.

(a) Find the MLE $\widehat{\theta}$ of $θ$.

(b) Find constants c and d (possibly depending on n) such that, c+d$\widehat{\theta}$ is unbiased for $θ$.

The likelihood of $θ$ being greater than $X_1$ is zero, but what is the likelihood of it being lower $X_1$. Is it the $θ^{-n}$ (same as when we try to estimate the higher limit of the Uniform Distribution.

I understand that $θ$ should be the minimum of the sample and 1 should be the max of the sample (ideally), but how to derive it mathematically?

$\endgroup$
2
1
$\begingroup$

The likelihood of a sample $\mathbf x=(x_1,\ldots,x_n)$ in $(0,1)$ is $$ L(\theta\mid\mathbf x)=\prod_{k=1}^n\frac{\mathbf 1_{[\theta,1]}(x_k)}{1-\theta}=\frac{\mathbf 1_{\theta\leqslant\min\mathbf x}}{(1-\theta)^n}. $$ Surely you can deduce which value $\hat\theta(\mathbf x)$ of $\theta$ maximizes this.

$\endgroup$

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.