The field of class prediction with microarray data has seen a lot of research activity in recent years. Owing to this effort, gene expression profiling is getting more and more established in clinical practice. The most prominent applications lie within cancer research: microarrays are often used to support exact phenotyping in early stages of the disease, which potentially allows for tailored treatment and better cure rates.

This chapter contains an easy-to-follow generic recipe for class prediction with R. It starts by explaining how to retrieve data and terminates by showing how to summarize class prediction results. The main focus lies on learning classification rules, which, since it usually requires user interaction, is the most complex and laborious step in a class prediction analysis.