|
Activity Number:
|
371
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #303780 |
|
Title:
|
Generalized Functional Latent Feature Models in Colon Carcinogenesis Data
|
|
Author(s):
|
Yehua Li*+ and Naisyin Wang and Raymond J. Carroll
|
|
Companies:
|
The University of Georgia and Texas A&M University and Texas A&M University
|
|
Address:
|
204 Statistics Building, 101 Cedar Street, Athens, GA, 30602,
|
|
Keywords:
|
Asymptotics ; functional data ; kernel smoothing ; principal components ; single-index model
|
|
Abstract:
|
p27 is an important biomarker in colon cancer studies. p27 measurements within the colon crypts are indexed by the relative cell depth, and therefore naturally considered as functional data. Our goal is to study the relationship between cell apoptosis and the p27 level. On the other hand, many environmental variables also play an important role in this study. We propose a new class of functional regression models, in which the coefficient function depends on a single-index of the environmental variables. The interaction between the functional predictor and the environmental variables is modeled nonparametrically. We propose a MAVE-like estimation procedure, and the estimator of the parametric components of the model is shown to be root-n consistent and asymptotically normal. Simulation studies and a real application are presented for illustration of the method.
|
