Activity Number:
|
472
- Junior Research in Bayesian Nonparametric Modeling of Complex or Unknown Populations
|
Type:
|
Topic Contributed
|
Date/Time:
|
Wednesday, August 1, 2018 : 8:30 AM to 10:20 AM
|
Sponsor:
|
International Society for Bayesian Analysis (ISBA)
|
Abstract #329719
|
|
Title:
|
Hierarchical Infinite Latent Factor Models
|
Author(s):
|
Elizabeth Lorenzi* and Ricardo Henao and Katherine Heller
|
Companies:
|
Duke University and Duke University and Duke University
|
Keywords:
|
hierarchical modeling;
nonparametrics;
latent factor model
|
Abstract:
|
We develop a hierarchical Bayesian latent factor model to appropriately account for the covariance structure across subpopulations in data. We propose a Hierarchical Dirichlet Process shrinkage prior on the loadings matrix that flexibly captures the underlying structure of our data across subpopulations while sharing information to improve inference and prediction. Additionally, the stick-breaking construction of the prior allows for each subpopulation to utilize different subsets of the factor space and select the number of factors needed to best explain the variation. Theoretical results are provided to show support of the prior. We focus on the use of the model for prediction and show that our method improves on prediction in individual populations compared to other factor and regression models.
|
Authors who are presenting talks have a * after their name.