[Forum SIS] Bocconi DEC Seminar: Peter Orbanz - October 7th

[Anna Simoni]

Dear colleagues, 



The Department of Decision Sciences (DEC) of Bocconi University is pleased to invite you to the seminar: 
================================================================== 
DEC Seminar 
Università Bocconi, 
Room 3-E4-SR03 
Via Rontgen 1 - 3rd floor 
Time: 12:30pm 


The DEC seminar schedule is available at http://www.unibocconi.eu/statseminar 

Thursday, October 7th 

Peter Orbanz 
( University of Cambridge ) 

"Conjugate Projective Limits" 

Abstract : 

Bayesian nonparametric models can be regarded as Bayesian models on infinite-dimensional 
spaces. These infinite-dimensional distributions can be constructed from finite-dimensional 
ones using the projective limit approach familiar from stochastic process theory. I will 
discuss how this approach can be generalized from probability measures to conditional 
probabilities. The generalization results in a projective limit calculus applicable 
to the construction and representation of virtually any nonparametric Bayesian model. 
Intuitively, rather than constructing a stochastic process from finite-dimensional 
probability measures, we construct a nonparametric Bayesian model from finite-dimensional 
Bayes equations. Specific properties of the finite-dimensional models -- such as posterior 
updates, sufficient statistics and posterior conjugacy -- carry over to the infinite-dimensional 
case. I intend to sketch a few construction examples and, time permitting, will discuss 
in how far the constructions are generic and what mathematical complications we have to expect. 

============================================================================== 

Sincerely, 
Anna Simoni 



--------------------------- 
Anna Simoni 
Assistant Professor 
Department of Decision Sciences 
Università Bocconi 
via Roentgen, 1 
20136 Milano - Italy 
Email: anna.simoni a unibocconi.it 
Webpage: http://faculty.unibocconi.eu/annasimoni/ 
-------------- parte successiva --------------
Un allegato HTML è stato rimosso...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20101005/c35caf84/attachment-0001.html>