News for hackers. Stuff that matters.
Re:A sane introduction to maximum likelihood estimation and ... by nerdponx on Thursday January 03, 2019 @02:35PM (#18818423)
I never suggested a definition of MLE. You need independence to use the "L = Π P(y | θ)" formulation, full stop.
Re:A sane introduction to maximum likelihood estimation and ... by conjectures on Thursday January 03, 2019 @05:30PM (#18819977)
Yes, you do need independence to assume the likelihood factorises. You do not need independence to find a MLE/MAP.
Re:A sane introduction to maximum likelihood estimation and ... by stilley2 on Thursday January 03, 2019 @02:49PM (#18818571)
True. But that form is a convenience, not a requirement.