Activity Number:
|
501
- Bayesian Penalized Likelihood Methods for Gaussian Graphical Models
|
Type:
|
Invited
|
Date/Time:
|
Thursday, August 11, 2022 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Section on Bayesian Statistical Science
|
Abstract #320556
|
|
Title:
|
New Directions in Bayesian Shrinkage for Structure Learning
|
Author(s):
|
Ksheera Sagar K. N. and Sayantan Banerjee and Jyotishka Datta* and Anindya Bhadra
|
Companies:
|
Purdue University and Indian Institute of Management Indore and Virginia Tech and Purdue University
|
Keywords:
|
graphical model;
non-convex optimization;
sparsity;
posterior contraction;
shrinkage prior;
structure learning
|
Abstract:
|
High-dimensional data with complex structure have become pervasive in many areas of science and engineering. The problem of precision matrix estimation in a multivariate Gaussian model is one such methodological problem, fundamental for network estimation. Although there exist both Bayesian and frequentist approaches, it is difficult to obtain good Bayesian and frequentist properties under the same prior-penalty dual, complicating justification. In this talk, I will briefly review recent developments in precision matrix estimation using global-local shrinkage priors and propose possible solutions that lead to a prior-penalty dual that offers fully Bayesian uncertainty quantification as well as computationally efficient penalized likelihood estimators.
|
Authors who are presenting talks have a * after their name.