abstract
- Support Vector Machines (SVM) have shown excellent generalization power in classification problems. However, on skewed data-sets, SVM learns a biased model that affects the classifier performance, which is severely damaged when the unbalanced ratio is ver