Enhanced support vector machine using parallel particle swarm optimization

Proper parameter settings of support vector machine (SVM) and feature selection are of great importance to its efficiency and accuracy. In this paper, we propose a parallel adaptive particle swarm optimization algorithm to simultaneously perform the parameter optimization and feature selection for SVM, termed PTVPSO-SVM. It is implemented in an efficient parallel environment using PVM (Parallel Virtual Machine). In the proposed method, a weighted function is adopted to design the objective function of PSO, which takes into account the average accuracy rates (Acc), the number of support vectors (SVs) and the selected features simultaneously. The adaptive control parameters including the time varying acceleration coefficients (TVAC) and inertia weight (TVIW) are employed to efficiently control the local and global search in PSO and mutation operators are introduced to overcome the problem of the premature convergence of PSO algorithm. The experimental results clearly confirm the superiority of the proposed method over the other two reference methods on several real world datasets. It also reveals that the PTVPSO-SVM can not only obtain much more appropriate model parameters, discriminative feature subset as well as smaller sets of SVs but also significantly reduce the computational time, giving high predictive accuracy.