e1df0d58c5f51410VgnVCM100000c2b1d38dRCRDapproved/UMICH/stats/Home/News & Events/Statistics SeminarJayaram Sethuraman###@###(Fri, 4 Oct 2013)Jayaram Sethuraman###@###(Fri, 4 Oct 2013)411 West HallA personal history of the constructive definition of the Dirichlet processstats1380900600000138090060000011:30 AM<p>Following standard parametric Bayes applications, there was a<br> growing need for nonparametric Bayes analysis which seemed to<br> pose some difficulties. These difficulties were removed by the<br> the first definitions of Dirichlet processes by Ferguson,<br> Blackwell and MacQueen. Dirichlet processes became famous after<br> some easy applications in standard nonparametric Bayes<br> problems. However, the proofs looked complicated and<br> mystifying. In spite of this many salient properties of Dirichlet<br> processes were being established.</p> <p>The Sethuraman constructive definition of a Dirichlet process<br> simplified proving such properties and establishing further<br> properties. Computational Bayes analysis took off in parametric<br> problems, thanks to MCMC. The importance of the constructive<br> definition for computational nonparametric Bayes analysis was<br> recognized immediately and it opened the flood gates for a<br> multitude of diverse applications.</p> <p>This talk will traverse through some of these developments.</p>Njkmcdonjkmcdon1380286560933b1df0d58c5f51410VgnVCM100000c2b1d38d____once11112newnewJayaram Sethuraman