Abstract
Several bio-inspired evolutionary based feature selection algorithms for microarray data classification have been proposed in the literature and show a good performance. In this research we proposed a wrapper feature selection algorithm for classifying cancer microarray gene expression profile that uses FireFly algorithm along with SVM classifier named FF-SVM. Support vector machine SVM classifier with leave-one-out cross validation LOOCV are used to measure the classification accuracy for the selected gene subset. Five benchmark microarray datasets of binary and multi class are used to evaluate FF-SVM algorithm. To validate the result of the proposed algorithm we compare it with other related state-of-the-art algorithms. The experiment proves that the FF-SVM show high classification accuracy using small number of selected genes.