Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

$X_1,\ldots,X_n$ $n$ sampled are random variable $f(x|\theta)=(-1)\theta \ln\theta 1_{x>\theta}$ where $1$ is the indicator function. What is the log likelihood function? It seems easy to multiply but i am not sure how to handle the term $(-1)^n$ and $1^n{x>\theta}$

share|improve this question
    
Some editing is needed to make it clear exactly what $f(x|\theta)$ is. –  Dilip Sarwate Mar 16 '12 at 1:06
add comment

1 Answer 1

$1_{x\gt \theta}$ means there is only a non-zero density or likelihood when $X_i$ is greater than $\theta$. So for the whole sample you require all the $X_i$ to be greater than $\theta$ or, in other words, $\theta \lt \min_i \{X_i\}$.

The $(-1)$ could be there because $\ln \theta$ would be negative if $0 \lt \theta \lt 1$, but a probability density must be non-negative.

I suspect there is an error in what you have written for the conditional density or at least something is missing, as integrating it over $x$ does not obviously give $1$. Something like

$$f(x|\theta) = \frac{(-1)}{x \ln \theta} \; 1_{\theta \lt x \lt 1}$$

might be what you intended.

share|improve this answer
    
But i think the existence of (-1) is enough to show that no log likelihood function exist right?? –  Mathematics Mar 16 '12 at 3:14
    
No. If $0 \lt \theta \lt 1$ then $-\log(\theta)$ is positive and so might be a likelihood and so $\log( -\log(\theta))$ is real and so might be a log-likelihood. –  Henry Mar 16 '12 at 7:13
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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