TABLE OF CONTENTS Gene selection is an important issue in microarray data analysis and has critical implications for the discovery of genes related to serious diseases. This paper proposes a Fisher optimization model in gene selection and uses Fisher linear discriminant in classification. Experiment result in public data has demonstrated validity of this method.
With development of the human genome project (HGP), a large number of genes have been found and located, and the study on the functions of genes has become an important issue in the post-genome project [ [1]]. With the rapid development of microarray technology [ [2]] and the great emergence of functional genome data in recent years, microarray data analysis has become an important area of study in bioinformatics. This has also brought about enormous challenge as well opportunities for pattern recognition and machine learning. In the perspective of pattern recognition and machine learning, microarray data analysis mainly covers gene expression pattern clustering, supervised learning, classification, gene selection, etc. [ [3], [4], [5], [6]]. An extensive concern and research have been given to these areas. Through discrimination of genes and their functions, mankind will find new ways to take precautions and finally cure serious diseases such as cancer and heart disease.
A human being has around 30 thousand genes [ [7]]. However, only a very...
