|
Activity Number:
|
510
- Recent Development in Semiparametric and Nonparametric Methods
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, July 31, 2019 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
International Chinese Statistical Association
|
|
Abstract #307348
|
|
|
Title:
|
Bayesian Penalized Spline Estimation for Generalized Partially Linear Single Index Models Using JAGS
|
|
Author(s):
|
Zhaohu(Jonathan) Fan* and Yan Yu
|
|
Companies:
|
University of Cincinnati and University of Cincinnati
|
|
Keywords:
|
Generalized Partially Linear Single Index Models ;
Penalized Spline;
Bayesian
|
|
Abstract:
|
Generalized partially linear single-index model (GPLSIM) is a useful tool for high dimensional nonparametric regression for responses from a general exponential family. Penalized splines (P-splines) have shown promising successes, especially in computation, to fit the univariate unknown function for GPLSIM. We present a fully Bayesian framework for GPLSIM with P-splines using a readily available R package JAGS. We consider spline coefficients as random effects in generalized linear mixed models that can be viewed to include penalized likelihood models as special cases. We demonstrate the proposed fully Bayesian approach through simulation studies and apply it to two real data examples.
|
Authors who are presenting talks have a * after their name.
