Maximizing the product of the priors is NP-complete.
Maximizing the sum of the posteriors given the posteriors is NP-complete (though often easier since constraints are accounted for).
Finding the posteriors is PP-complete (decision version of #P).
