Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 26 - Imaging Speed Session
Type: Contributed
Date/Time: Sunday, August 8, 2021 : 1:30 PM to 3:20 PM
Sponsor: Section on Statistics in Imaging
Abstract #318488
Title: Clustering Using Probabilistic JIVE with Gaussian Mixtures
Author(s): Ganzhong Tian* and John Hanfelt and Raphiel Murden and Deqiang Qiu and Benjamin Risk
Companies: Emory University and Emory University and Emory University and Emory University and Emory University
Keywords: Clustering; JIVE; Gaussian Mixture; PCA; Brain Imaging; Data Integration

A common goal in data integration studies is to identify subgroups. Recently, JIVE (joint and individual variation explained) has been proposed as a method to extract shared (joint) and unique (individual) information from each dataset, and k-means is applied after extraction of joint and individual scores. We present a probabilistic JIVE model with mixture of Gaussians (ProJIVE-mix), which enables joint probabilistic clustering of subjects with multiple data sources. Our simulations demonstrate improvement over concatenating datasets. We also apply our method to clustering subjects with fMRI brain imaging and CSF biomarker measurements in the Alzheimer's Disease Neuroimaging Initiative.

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

Back to the full JSM 2021 program