2v-SSPC: A new classification method for class-imbalanced data

Using data sets that contain very few instances of the positive class usually produces the biased classifier and the predictive accuracy over the positive class (usually the more important class) is worse than that over the negative class. A classification method for imbalance data is proposed. This a obtains method maximum separation ratio to separate two classes instances via a single hypersphere and also provides the facility to control the upper bounds of two classes error rates respectively with two parameters. As such, the performance of classification and prediction of imbalanced data sets can be improved, and the range of selection of parameters can be greatly narrowed. Using area under the ROC curve as performance measurement, experimental results on UCI data sets show the method's effectiveness.