Model selection of RBF kernel for C-SVM based on genetic algorithm and multithreading

Abstract

Generalization performance of support vector machines depends on optimal selection of parameter values. But training the best parameters for C-Support Vector Machines (C-SVM) classifier with RBF kernel is time-consuming. We can hardly finish training process for large data sets with traditional methods. Multithreading as a widespread programming and execution model allows multiple threads to exist within the context of a single process, which has been widely applied in data processing and analyzing. In this paper, we studied how to adopt genetic algorithm and multithreading model to complete optimal model selection of C-SVM classifier with RBF kernel. This new approach not only chooses global parameters, but also saves training time based on parallel computing process. Experimental results show the efficiency and feasibility of new approach.