Classification optimization using PSO-SSO based support vector machine

Classification is one of the widely used technique for data mining of dataset and is done using soft computing approach. Here a novel method called SSO-ELS is proposed for classification of datasets. In this method there is hybridization of Simplified Swarm Optimization (SSO) with ELS (Exchange Local Search), Particle Swarm Optimization (PSO) and Support Vector Machines(SVM) approach. This is done to resolve the issue of selection of hyper parameters in SVM. The selection of hyper parameters in SVM plays a crucial rule which is done by the PSO-SSO approach. This approach has two phases: In first phase best initial parameters of SVM are calculated using SSO with ELS approach and then the best parameters are fed into SVM using PSO in second phase. Brief review of classification methods is discussed. Experiments on UCI datasets indicate that the proposed SSO-PSO-SVM achieves better results than CS-PSO-SVM with respect to classification accuracy and F-measure.