Gene Selection with Rough Sets for the Molecular Diagnosing of Tumor Based on Support Vector Machines

Abstract

The development of microarray technology has motivated interest of its use in clinical diagnosis of tumor and drug discovery. However the accurate classification of tumor by selecting the tumor-related genes from thousands of genes is a difficulty task due to the large number of redundant genes. Therefore, we propose a novel hybrid approach which combines rough set theory with support vector machines to further improve the classification performance of gene expression data. Our approach is assessed on two well-known tumor datasets, and experiments indicate that gene selection based on the rough set theory is effective because most of the selected genes are relevant to tumor using rough set attribute reduction, and support vector machines classifier has a better performance on the selected informative genes.