|
Activity Number:
|
387
- Software
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 12, 2021 : 12:00 PM to 1:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
|
Abstract #318838
|
|
|
Title:
|
Reduced Rank Estimation in Mixtures of Multivariate Linear Regression
|
|
Author(s):
|
Suyeon Kang* and Kun Chen and Weixin Yao
|
|
Companies:
|
University of California, Riverside and University of Connecticut and University of California, Riverside
|
|
Keywords:
|
Reduced rank estimation;
Multivariate linear regression;
Mixtures of regression models
|
|
Abstract:
|
Motivated by the idea of reduced rank estimation originally developed for non-mixture cases, we here study reduced rank mixtures of multivariate response regression models to provide more parsimonious and interpretable models. The proposed estimator simultaneously take take into account the joint structure of the multivariate response and population heterogeneity. We show the complete derivation of iterative algorithms that perform parameter estimation in mixtures of multivariate response regression models with and without the reduced rank framework. Via the proposed paradigm, we have some desired features such as monotonicity of the penalized likelihood sequence. The consistency of the proposed estimators is established. The performances of the proposed reduced rank methods are evaluated through simulation studies and real data analysis.
|
Authors who are presenting talks have a * after their name.
