hi, i think i have found a gap in my knowledge. i have three questions on this topic can you please help?
1) i thought that we always did Z*mle+(1-Z)*prior mean
but now im thinking maybe prior mean isnt right - is it whatever bayesian type we are splitting ie if splitting under all or nothing would it be Z*mle+(1-Z)*prior estimate under all or nothing
2)do we assume the mle is x/n unless we are told to find the mle in a previous part. i have never found a question where the mle is not x/n without a earlier part making us find it?
3) i have found that we do not do the product part of the likelihood here sum as sumx and power n. is it true we do not treat the likelihood as n likelihoods but just as singular for these quetsions?
thank you
Last edited: Apr 18, 2023