|
Activity Number:
|
87
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 31, 2016 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
|
Abstract #320391
|
|
|
Title:
|
Model Fitting and Optimal Design for a Class of Binary Response Models
|
|
Author(s):
|
Subir Ghosh* and Hans Nyquist
|
|
Companies:
|
University of California at Riverside and Stockholm University
|
|
Keywords:
|
Binary Response ;
Efficiency Comparison ;
Estimating Equations ;
Maximum Likelihood ;
Model Discrimination ;
Odds Ratio
|
|
Abstract:
|
A class of binary response models is considered for describing the data on a response variable having two possible outcomes and q explanatory variables when the odds ratios on the response are a linear function of the explanatory variables. The models provide the closed form solutions of the maximum likelihood estimating equations for the parameter estimation under a Bernoulli setup. A data example is presented to demonstrate the better goodness of fit of a model within this class in comparison with the logit, probit, and complimentary log-log models. The design conditions are given and locally optimal designs are presented for some special cases under the optimality criterion functions. The issues in the model discrimination are discussed with an illustrative example.
|
Authors who are presenting talks have a * after their name.
