Gene selection for cancer classification using the combination of SVM-RFE and GA

Xiaobo Li

COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(4) 262-267

Department of Computer Science and Technology, College of Engineering, Lishui University, Lishui 323000, China

Gene selection is a key research issue in molecular cancer classification and identification of cancer biomarkers using microarray data. Support vector machine recursive feature elimination (SVM-RFE) is a well known algorithm for this purpose. In this study, a novel gene selection algorithm is proposed to enhance the SVM-RFE method. The proposed approach is designed to use the combination of SVM-RFE and genetic algorithm (GA). The performance of the proposed model is validated on a binary and a multicategory microarray gene expression datasets. The results show that the proposed gene selection method is able to elevate the performance of SVM-RFE, which extracts much less number of informative genes and achieves highest classification accuracy.