Represent uncertainty with prior distributions, likelihoods, and posterior distributions. Build conjugate models and interpret credible intervals, posterior probabilities, and predictive distributions.