Inference procedures for multivariate and functional data : applications to pharmaceutical statistics and remote sensing data analysis.

This dissertation is comprised of three research papers which span topics in multivariate data analysis and functional data analysis. We first provide a brief introduction to relevant background information in chapter one. In the second chapter, we propose a novel approach for estimating the common mean of a multivariate normal distribution, which is shown to be especially advantageous when the sample size is small relative to the data-dimension. We then extend this approach to the two-sample case in chapter three, where the work is motivated by an application in pharmaceutical statistics. In chapter four we present a two-stage variable selection technique designed for functional data, motivated largely by an application in remote sensing data analysis. We conclude with a brief summary and discussion in chapter five.