Prior Distribution

Abstract In Bayesian inference, the unknown parameter is given a prior distribution, and Bayes' theorem combines this with the likelihood from the observed data, to give the posterior distribution. The prior may be informative, subjective, or have a standard, reference form (usually representing vague knowledge). In hierarchical models, the prior distribution may have its own parameters that themselves have priors.