To answer Kristin's question about the samples, I looked at the introduction to section five. Sample were taken from 78 leukemia patents, and the task was to classify the leukemia as type I or type II. Apparently this classification can be made by the symptoms at some point in the disease, but the task is to identify the type earlier based on gene expression. Gene expression is a measure of whether a certian segment of DNA is currently being used in the process of making proteins. Most of a cell's DNA is dormant at any given time. I do not know how they check for gene expression, but I suspect it is measured by the RNA in the cells. RNA is a modified version of DNA that is directly involved in protein construction.

Anyone in bioinformatics care to confirm/deny my analysis?