Re: MLE

On May 23, 1:56 pm, Ray Koopman <koop...@xxxxxx> wrote:
On May 23, 4:41 am, sagar <ariji...@xxxxxxxxx> wrote:

Can anybody give me an example where MLE works but OLS does not.

Basically, any model in which the errors are not approximately
independent homoscedastic normal.

Here's a nice tutorial on MLE -->
http://www.psy.vanderbilt.edu/faculty/palmeri/P351-modeling/readings/myung-tutorial-mle.pdf

HTH

Ryan
.

Loading