Archive for Athens University of Economics and Business

Greek variations on power-expected-posterior priors

Posted in Books, Statistics, University life with tags , , , , , , on October 5, 2016 by xi'an

Dimitris Fouskakis, Ioannis Ntzoufras and Konstantinos Perrakis, from Athens, have just arXived a paper on power-expected-posterior priors. Just like the power prior and the expected-posterior prior, this approach aims at avoiding improper priors by the use of imaginary data, which distribution is itself the marginal against another prior. (In the papers I wrote on that topic with Juan Antonio Cano and Diego Salmerón, we used MCMC to figure out a fixed point for such priors.)

The current paper (which I only perused) studies properties of two versions of power-expected-posterior priors proposed in an earlier paper by the same authors. For the normal linear model. Using a posterior derived from an unormalised powered likelihood either (DR) integrated in the imaginary data against the prior predictive distribution of the reference model based on the powered likelihood, or (CR) integrated in the imaginary data against the prior predictive distribution of the reference model based on the actual likelihood. The baseline model being the G-prior with g=n². Both versions lead to a marginal likelihood that is similar to BIC and hence consistent. The DR version coincides with the original power-expected-posterior prior in the linear case. The CR version involves a change of covariance matrix. All in all, the CR version tends to favour less complex models, but is less parsimonious as a variable selection tool, which sounds a wee bit contradictory. Overall, I thus feel (possibly incorrectly) that the paper is more an appendix to the earlier paper than a paper in itself as I do not get in the end a clear impression of which method should be preferred.

MCMC convergence in Athens (back)

Posted in Statistics, Travel, University life with tags , , , on September 15, 2009 by xi'an

athens1While I remain somehow disappointed by my trip to the Acropole, I bring back ideas and directions from the MCMC workshop. As indicated earlier, this was a very small meeting with “only” eight people, in the same spirit as the “kitchen workshops” we used to run during the blessed days of the HSSS program, more than ten years ago… I am thus quite thankful to Petros Dellaportas and to Ioannis Kontoyiannis for creating this great opportunity for intense if relaxed exchanges on Monte Carlo methods!

Posters at auebPosters at aueb

PASO poster (C.) Sean MeynAmong those directions, I found a renewed interest in doing  (Dirichlet?) non-parametrics in Markov settings, following a talk by Persi Diaconis,  and in studying further control variates in connection with past and recent Rao-Blackwellisation papers of mine, following talks by Petros Dellaportas and Susan Holmes, while discussions with Christophe Andrieu opened new vistas for ABC approximation. Sean Meyn’s talk contained an interesting example of a very slowly converging Markov chain, with a neat presentation of the corresponding reasons at the large deviation level. Both Christophe Andrieu and Eric Moulines gave talks about adaptive schemes that showed me how fast things were moving in this area. Here are the slides of the talk I gave, mostly recycled from earlier talks already posted:

The pictures posted here were taken by Sean Meyn and myself inside the university (AUEB) building, where a fierce and political campaign for the students’ union seemed to be raging! I particularly like the (Leninist) PAΣ-EAAK poster of the May 68 students with a code-bared étudiant(e) from a more recent French demonstration…

KKE tag in aueb (C.) Sean MeynAUEB