Paper Info
Title
A Compound Optimality Criterion Ford-Efficient And Separation-Robust Designs For The Logistic Regression Model
Abstract
TheDMP-criterion is proposed to generate optimal designs for the logistic regression model with reduced separation probabilities. This compound criterion has two components: (a) theD-efficiency of the candidate design and (b) a penalty term that captures the average distance of the candidate design's support points from the region of maximum prediction variance (MPV). ADMP-optimal design maximizes theDMP-criterion. The aim is to obtain compromise experimental designs with highD-efficiencies that are more robust to separation than aD-optimal design of equal size. This paper presents theDMP-criterion and demonstrates examples of its potential use as a means of mitigating separation in the design phase of a binary response experiment. For the examples presented, the localDMP-optimal designs offer a 20-30% reduction in separation probability over the localD-optimal designs while maintainingD-efficiencies over 93%. A robust design methodology is also demonstrated, where a robustDMP-optimal design is compared to a BayesianD-optimal design and shown to have comparableD-efficiencies across a range of randomly drawn parameter values while offering a mean reduction in separation probability of 23.9%.
Year
DOI
Venue
2021
10.1002/qre.2768
QUALITY AND RELIABILITY ENGINEERING INTERNATIONAL
Keywords
DocType
Volume
coordinate exchange, D-optimal, experimental design, logistic regression model, nonlinear, optimal design, separation
Journal
37
Issue
ISSN
Citations 
7
0748-8017
0
PageRank 
References 
Authors
0.34
0
3
Name
Order
Citations
PageRank
Anson R. Park100.34
Michelle V. Mancenido200.34
Douglas C. Montgomery310624.05