Scientific Computing Research Environments for the Mathematical Sciences (SCREMS)

University Of Connecticut, Storrs CT

Investigators

Abstract

We propose to purchase two Dell PowerEdge 1955 Gig E High-Performance Clusters and IMSL F90 and C numerical libraries as well as other software. The equipment and software will be dedicated to the support of research in a variety of projects, including statistical analysis for neuronal electrophysiological data, Bayesian structural equations modeling with applications to Veterans Health Administration survey data, statistical analysis of patterns of genetic variation in hierarchically structured populations, the effect of missing values on inference, and analysis of high-frequency financial data. Through these projects, we aim to make contributions to statistical methodology and computing, complex statistical modeling, software development and mathematical statistics. The proposed equipment will also greatly enhance the research and education infrastructure of the Department of Statistics at University of Connecticut at all levels. The research supported by the proposed high-performance equipment covers areas spanning from medical science, genetics, brain science, finance to health care. Discoveries in these areas constantly have far reaching impact to society, economy, science and technology. However, researchers in these areas are increasingly facing challenges of analyzing large and complex datasets. The proposed equipment will provide us with invaluable capacity to address the challenges, hence significantly facilitating our efforts to make scientific and technological breakthroughs.