High Performance Statistical Computing

This past semester I’ve been working three separate big data projects while on sabbatical. This talk will highlight the activities and some preliminary outcomes. The projects are: Lake Michigan Wind Assessment, data quality and modeling with dozens of variables collected once a second for roughly 9 months each during 2012 and 2013. Meijer Marketing Analytics, working with trillions of transactions and hundreds of variables using SAS and SQL on their new Teradata high performance computational facilities. Empirical Spectral Test, collaborative research with WMU Computer Science and Statistics faculty using multidimensional spatial Fast Fourier Transforms in R on our analytics server and at the WMU High Performance Computational Science Laboratory