Online Program Home
My Program

Abstract Details

Activity Number: 253 - Contributed Poster Presentations: Section on Statistical Computing
Type: Contributed
Date/Time: Monday, July 30, 2018 : 2:00 PM to 3:50 PM
Sponsor: Section on Statistical Computing
Abstract #328746
Title: An Efficient Algorithm for Outlier Detection in Linear Mixed Model
Author(s): Tzu-Ying Liu* and Hui Jiang
Companies: University of Michigan and University of Michigan
Keywords: linear mixed model; outlier detection; penalized maximum likelihood estimation

The best linear unbiased predictor (BLUP) from a linear mixed model gives moderated and efficient predictions for clustered or correlated data. However, when there are outliers among individual observations and random effects, the prediction accuracy could be hampered. Also, these outliers may suggest aberrant data generating mechanism that warrants further investigation. Motivated by an RNA-Seq dataset, we illustrated a case where there are outliers among individual observations and random effects. We also showed the connection of these outliers to the underlying issues that need to be identified. As the number of genes is substantial, we proposed a scalable algorithm to achieve simultaneous linear mixed effect model estimation and automated outlier detection.

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

Back to the full JSM 2018 program