Activity Number:
|
283
- Deep Learning Methods
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 9, 2022 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
Abstract #322631
|
|
Title:
|
Classification of Longitudinal Data with Irregularly Spaced Intervals: A Comparison Between Posthoc Mixture Modeling of BLUPs from Mixed-Effects Model and Deep Clustering Methods
|
Author(s):
|
Md Jobayer Hossain* and Ben Leiby
|
Companies:
|
Nemours Children's Health and Thomas Jefferson University
|
Keywords:
|
Classification;
longitudinal;
Mixture-based mixed-effects models;
posthoc mixture modeling;
deep learning methods
|
Abstract:
|
Classification of longitudinal trajectories in naturally occurring data is an area of growing interest in learning healthcare systems and precision medical decisions. Mixture-based mixed-effects models implemented in R and Mplus are commonly used methods for clustering longitudinal data with irregularly spaced measurements. The approaches become computationally complex with increased sample size and level of unbalancedness. In our recent large data applications and simulations, posthoc mixture modeling of BLUPs (PMMB) from mixed-effects models appeared to perform better than the two above existing mixture-based mixed-effects models. The method is flexible and is less affected by increased sample size. In recent developments, deep neural networks have been used in clustering large longitudinal datasets. In this study, we will compare the performance of deep learning methods and PMMB in clustering longitudinal data with irregularly spaced intervals.
|
Authors who are presenting talks have a * after their name.