Resumen
In recent years, support vector machine (SVM) based on empirical risk minimization is supervised learning model which has been successfully used in the classification and regression. The standard soft-margin SVM trains a classifier by solving an optimization problem to decide which instances of the training data set are support vectors. However, in many real applications it is imperative to perform feature selection to detect which features are actually relevant. In order to further improve the performance, we propose the adaptive condensed instances (ACI) strategy based on the hybrid particle swarm optimization (HPSO) algorithm for the SVM classifier design. The basic idea of the proposed method is to adopt HPSO to simultaneously optimize the ACI and SVM kernel parameters for the classification accuracy enhancement. The numerical experiments on several UCI benchmark datasets are conducted to find the optimal parameters for building the SVM model. Experiment results show that the proposed framework can achieve better performance than other published methods in literature and provide a simple but subtle strategy to effectively improve the classification accuracy for SVM classifier.