PCA-based gene selection for cancer classification

Gene expression has a vast area of application in diagnosis of disease in the medical field. In this gene expression data, the quantity of genetic factor involved is very large (in ten thousand) associated to the number of samples which is very few in cancer classification. This large number of genes are inefficient as the classifier accuracy decreases. A novel method is to reduce the size of the dataset to contain only most prominent genes that can be used to increase the classification accuracy. Our method use PCA based feature selection and SVM-RFE classification algorithm.