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.

In statistics one often hears the phrases "log likelihood" and "log error". Why is it natural to consider the logarithm of these guys? My wild guess is that the Gaussian is a pretty "normal" distribution, and it's natural to take the log of such a thing. But this is pretty sketchy.

A pointer to a reference would be great.

share|improve this question
    
Why does someone think this question is off-topic, I wonder? It would also be on topic on stats.SE, but that doesn't make it unwanted here. –  Henning Makholm Feb 5 '13 at 21:42
1  
it is like you wanna marry a beautiful girl and to get her is very difficult.. however her mother is easier to convince and you also know that the girl you love always listens to her mother. What is the best strategy? first to get the mothers mind and eventually you will get the girl.. other way around is difficult.. now what is log? –  Seyhmus Güngören Feb 5 '13 at 23:40
add comment

1 Answer 1

up vote 1 down vote accepted

Motivation: It's just easier to work with log-likelihood sometimes (e.g. if the likelihood is an exponential function).

Justification: Since log is an increasing function, whatever maximises log-likelihood also maximises likelihood.

Wikipedia on Log-likelihood

share|improve this answer
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.