The current molecular biology and systems biology is featured by the rapid accumulation of high-throughput genomics and proteomics data like microarray and mass spectrometry (MS) data. Typical applications of these high-throughput biological data include the use of microarray or MS data for the molecular classification of complex human diseases like cancers, and for discovering the genes/protein markers (biomarkers) underlying the classification. Such tasks have been intensively studied as a new application field of pattern recognition in recent years. Almost all existing methods for classification and feature selection have been tried, and a large number of new or revised methods have been proposed.