I’m wondering if I should invest the time to memorize MGFs for all distributions or resort to knowing how to derive them from principles?
Thanks!
As a general rule of thumb I would memorize the main ones like normal, gamma, poisson, and anything that is explicitly called out in the syllabus. For me personally, I know I can do the integrals and series to derive these, but I wouldn’t want to be spending precious exam time to do that. You should also know how to derive one from first principles in case you are given a more contrived distribution.