Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Optimal Supersaturated Design for Finite Sample Populations
Abstract
Numerous construction methodologies have been developed in literature to construct balanced supersaturated designs (SSD) from the unconstrained $2^p$ sample population. The first part of the thesis introduces introduces a row exchange algorithm construction methodology to construct optimal SSDs efficiently from a finite sample population using UE(s2) as the evaluation criteria. Computational results show the algorithm consistently produces optimal SSDs in a computationally efficient manner. The second part of the thesis introduces introduce an optimality criteria for supersaturated designs that is based on an underlying inference methodology, the asymptotically optimal confidence regions for high-dimensional data1, which has been shown to have good theoretical properties, as well as expected coverage on simulated data. The criteria is defined for confidence intervals of single components from a supersaturated design. Supersaturated designs with better criteria should produce shorter desparsified lasso confidence intervals while maintaining coverage. An optimal condition for single component confidence intervals is also derived.
Subject Area
Statistics
Recommended Citation
Luo, Binjie, "Optimal Supersaturated Design for Finite Sample Populations" (2019). ETD collection for University of Nebraska-Lincoln. AAI22616215.
https://digitalcommons.unl.edu/dissertations/AAI22616215