JSM 2013 Home
Online Program Home
My Program

Abstract Details

Activity Number: 52
Type: Invited
Date/Time: Sunday, August 4, 2013 : 4:00 PM to 5:50 PM
Sponsor: IMS
Abstract - #307164
Title: Sparse Non-Negative Tensor Factorizations for Large Contingency Tables
Author(s): Anirban Bhattacharya*+
Companies: Duke University
Keywords: Bayesian ; Contingency table ; High-dimensional ; Posterior convergence ; Sparsity ; Tensor decomposition

High-dimensional unordered categorical data appear in a staggering number of areas ranging from marketing applications, behavioral and social sciences, government surveys, web data, relational networks, genetic epidemiology among others. Such data can be organized into a massive dimensional contingency table that would typically be extremely sparse. Traditional approaches for contingency table analysis fail to scale up to moderate dimensions and alternative approaches based on non-negative tensor factorization hold much promise in this regard. With such motivation, we develop sparse tensor factorizations for multivariate categorical variables where the number of variables can be potentially much larger than the sample size. Theoretical properties of the proposed method are investigated and operating characteristics are examined through simulated and real data examples. We also provide a modification of our proposed method to allow for tests of association between groups of variables, with application in genetic epidemiology.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2013 program

2013 JSM Online Program Home

For information, contact jsm@amstat.org or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

ASA Meetings Department  •  732 North Washington Street, Alexandria, VA 22314  •  (703) 684-1221  •  meetings@amstat.org
Copyright © American Statistical Association.