6
$\begingroup$

I have been told that a maximum likelihood estimate (MLE) does not always actually exist. Why is this the case? It is clear that the MLE may not be unique, but there should always be a maximum, no?

  • 2
    Related:http://stats.stackexchange.com/questions/16758/is-there-always-a-maximizer-for-any-mle-problem2011-11-26

3 Answers 3