I am now familiar with the Bayesian thinking process of using a prior and then getting the posterior once we observe data using the prior.

I read the following statements which I am trying to get my intuition but unable so far. Any intuitive explanation will help.

A conjugate prior may always not lead to a better MAP estimate. Also, it is not clear what is a "better" MAP estimate

Why does a conjugate prior always not lead to a better MAP? I thought if you have a conjugate prior it is always easy to maximise the posterior integral.

Another downside of a choosing a conjugate distribution for the prior is that for some problems the conjugate may be inadequate.

In the above context what does inadequate actually mean?

Thanks a lot. Much appreciated.


1 Answer 1


Conjugate priors may not be the best model choice if they are chosen for convenience (i.e., ease of deriving the posterior). Your priority -- ideally -- should be to select a prior that best describes your belief. For example, suppose you have a Poisson likelihood and have a prior belief that the rate parameter is uniformly likely to be (1,7), then you should use a uniform prior. You should not use a Gamma(.) just because it would yield a Gamma posterior. This should answer both questions. If it doesn't, please let me know if you need any clarifications.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.