3 major components:
prior distribution of θ
likelihood of θ given the data
posterior distribution of θ