|
Activity Number:
|
6
|
|
Type:
|
Invited
|
|
Date/Time:
|
Sunday, August 3, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #300153 |
|
Title:
|
Modeling Sparse Generalized Longitudinal Observations with Latent Gaussian Processes
|
|
Author(s):
|
Fang Yao*+ and Hans G. Müller+ and Peter G. Hall
|
|
Companies:
|
University of Toronto and University of California, Davis and The University of Melbourne
|
|
Address:
|
Department of Statistics, Toronto, ON, M5S3G3, Toronto Department of Statistics, Davis, CA, 95616,
|
|
Keywords:
|
Binomial Data ; Functional Principal Component ; Functional Data Analysis ; Generalized Repeated Measurements ; Prediction ; Smoothing
|
|
Abstract:
|
In longitudinal data analysis one frequently encounters non-Gaussian data that are repeatedly collected for a sample of individuals over time. These observations could be binomial, Poisson or continuous. The timings of the repeated measurements are often sparse and irregular. We introduce a latent Gaussian process model for such data, establishing a connection to functional data analysis. We develop functional principal component (FPC) analysis for this situation and demonstrate the prediction of individual trajectories from sparse observations. This method can handle missing data and leads to predictions of the FPC scores. We illustrate these nonparametric methods with longitudinal data on primary biliary cirrhosis and show in simulations that they are competitive in competitive in comparisons with Generalized Estimating Equations and Generalized Linear Mixed Models (GLMM).
|
