{"id":"arxiv/2606.24519","url":"https://arxiv.org/abs/2606.24519","tags":["math.ST"],"score":0,"title":"Ferguson's Dirichlet Process Breakthrough: A Lasting Legacy","domain":"math.ST","source":"arxiv","authors":["Antonio Lijoi","Igor Pruenster","Junyi Zhang"],"reviews":0,"subject":"math.ST","summary":"Ferguson's 1973 introduction of the Dirichlet process marked a breakthrough in Bayesian nonparametric statistics. For the first time, a prior on the space of probability measures fulfilled two key desiderata: large support and analytical tractability. In this paper, we review three complementary constructions of the Dirichlet process, whose roots can be traced back to Ferguson: through finite-dimensional distributions, via normalization of a gamma process, and through predictive distributions. Each perspective not only deepens the understanding of the Dirichlet process but also provides a…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-06-23","reactions":{}}