Văn bản này: Efficient resampling methods for training support vector machines with imbalanced datasets