Support vector machine with parameter optimization by bare bones differential evolution

As one state-of-the-art pattern recognition method, support vector machine (SVM) has been successfully applied in many diverse regions. But the parameter optimization for SVM is a further ongoing research issue. The most used grid search method is time-consuming and the traditional global stochastic optimization techniques are parameter dependent. In this paper, bare bones differential evolution (BBDE) is used to tune the parameters of SVM. The BBDE is a new, almost parameter-free optimization algorithm that is a hybrid of the barebones particle swarm optimization (PSO) and differential evolution (DE). Some international standard data sets are used to evaluate the proposed algorithm. The experiment shows that BBDE has good performances to find the optimal parameters for SVM and is superior to some other methods.