In recent years, high-dimensional measurements have increasingly become available for diagnosis. The most familiar high-dimensional measures are gene expression arrays, but other settings include spectral, proteomic and metabolomic data.

A high-dimensional classifier challenges many of the standard methodologies used in diagnostics, and some issues and solutions will be outlined.

Functional data such as emission spectra are a special class of high-dimensional classifier whose continuity creates some modeling opportunities not seen in general high-dimensional settings. These will be outlined and illustrated in the context of cancer diagnosis.