Investigator:
Fred Warner
Start Date:
May, 2000
Description:
We will develop mathematical tools for the analysis of hyperspectral data. This will include, but not be limited to, spectroscopic data for the detection of tumors and biological contaminants. The high-dimensionality of the data necessitates the develoment of algorithms for compression of the data with various goals in mind. With this in mind, we will investigate techniques for the detection of local anomalies, feature extraction for detection and regression problems, and enhancements of traditional chemometric tools.