Abstract
Optimal designs are computer-generated experimental designs that provide an experimenter with an ‘optimal’ set of experimental trials. Historically, optimal experimental design has been limited to optimization with regards to a single criterion for a single response variable. Recent research by Burke et al. (2017) made it possible to create a dual response optimal designs for cases involving experiments with one continuous response and one binary response. The algorithm in Burke et al. (2017) provides a series of weighted optimal designs across a range of weights between the continuous and binary response cases. This thesis extends the work by Burke et al. (2017) in three ways. First, a new optimality criterion is developed in order to provide more stable algorithm results. Second, a method for selecting the weighted design that provides the best results for the continuous and binary cases is developed. Finally, a sensitivity analysis on the prior information required to generate the optimal designs in performed.
Library of Congress Subject Headings
Experimental design--Data processing; Optimal designs (Statistics); Bayesian statistical decision theory
Publication Date
4-26-2019
Document Type
Thesis
Student Type
Graduate
Degree Name
Industrial and Systems Engineering (MS)
Department, Program, or Center
Industrial and Systems Engineering (KGCOE)
Advisor
Rachel Silvestrini
Advisor/Committee Member
Katie McConky
Recommended Citation
Little, Rory W., "Advancing implementation conditions of optimal experimental design for dual response systems using one continuous response and one binary response" (2019). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/10017
Campus
RIT – Main Campus
Plan Codes
ISEE-MS
