To me, it will require a distributional assumption to the given data, otherwise we cannot define likelihood function. I feel this way, but I want to be sure if I am correct or not.
Does Maximum Likelihood Estimation require any distributional assumptions?
0
$\begingroup$
statistics
data-analysis
-
1yes it requires you to assume a distribution and estimate it's parameters to maximize the likelihood to obserse your dataset under the hypothesis that your data follows the assumed distribution – 2017-01-24
-
1Yes it requires a finite dimensional (finite parameters) distribution assumption but this can be as general as want. – 2017-01-24
-
0Thank you for the comments. Just one more thing that have come across, we can also use MLE for non-parametric with no distributional assumption, right? – 2017-01-24