Date of Award
Master of Science
Using semi-parametric models with Gaussian kernels, a variable selection process is proposed. Through simulations, the proposed method, which can select either continuous or discrete variables which may have complex interactions, is evaluated. The method is employed in a real data application to identify important CpG sites and single nucleotide polymorphisms (SNPs) in a set of genes such that DNA methylation of those sites and SNPs may have joint effects on allergic sensitization.
dissertation or thesis originally submitted to the local University of Memphis Electronic Theses & dissertation (ETD) Repository.
Terry, William G., "A Bayesian Semi-Parametric Approach to Variable Selection" (2015). Electronic Theses and Dissertations. 1120.