JSM 2015 Preliminary Program

Online Program Home
My Program

Abstract Details

Activity Number: 165
Type: Topic Contributed
Date/Time: Monday, August 10, 2015 : 10:30 AM to 12:20 PM
Sponsor: Section on Physical and Engineering Sciences
Abstract #315485 View Presentation
Title: Approximately Optimal Experimental Designs for Generalized Linear Mixed Models
Author(s): Xiaojian Xu* and Sanjoy Sinha
Companies: Brock University and Carleton College
Keywords: D-optimality ; longitudinal data ; mixed model ; approximately optimal design ; 'pseudo' Fisher Information ; robust design
Abstract:

We discuss optimal sequential designs for generalized linear mixed models (GLMMs) which have been broadly used in the analysis of longitudinal data or repeated measurements. Previous research has indicated that highly intensive computation is required for the construction of optimal designs for GLMMs mainly due to the integration involved in deriving the full Fisher information. In the present paper, we propose a method of constructing approximately optimal designs based on 'pseudo' Fisher information. In particular, we investigate the performance of the proposed designs through simulation studies. Our simulation results indicate the loss of efficiency from such approximation in optimal design construction process is minimal. Further, the resulting designs are robust against possible misspecification in the assumed random effects distribution. It is concluded that the proposed approach is of practical value for designing GLMM experiments.


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

Back to the full JSM 2015 program





For program information, contact the JSM Registration Department or phone (888) 231-3473.

For Professional Development information, contact the Education Department.

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

2015 JSM Online Program Home