Bayesian Nonparametric Inference: Dependence Structures and their Applications (17w5060)

Arriving in Oaxaca, Mexico Sunday, December 3 and departing Friday December 8, 2017


(Bocconi University)

(Universidad Nacional Autónoma de México)

(The University of Texas at Austin)


Despite the massive development of Bayesian Nonparametric (BNP) techniques in the last 40 years, there is an urgent need for new models and methodology that address practical issues that arise from diverse applied fields. Modern challenges have to do with the high-dimensionality of the data and with the complex dependence structures they feature: current methods and computational tools are not adequate to provide effective answers to a number of inferential problems that involve any of these two aspects. Indeed, traditional BNP models are not able to capture forms of dependence more general than exchangeability, which are, however, required in the analysis of several phenomena in epidemiology, genetics, medicine, finance, economics, etc.

Hence, the theme of the workshop will be the interplay between methodology and applications. A comprehensive environment, ranging from theoretical methodology to real-world applications of BNP models, is the key for the success of the proposed workshop.

The workshop will gather leading experts in BNP for a program that is designed to mainly focus attention on:
(a) methodological issues related to inference characterized by complex dependence structures;
(b) new computational tools that are needed for their actual implementation;
(c) successful applications in a wide spectrum of fields and new emerging methodological and computational challenges.

Therefore the workshop seeks to initiate collaborations among experts in theory and methods, researchers with complementary interests in applications and substantive scientific research questions, and foster interactions with scholars having strong computational skills.