The class of stick-breaking priors and their extensions are considered in classification and clustering problems in which the complexity, the number of possible models or clusters, can be either ...