多任务学习的不平衡SVM+算法

处理不平衡数据分类时，传统支持向量机技术（SVM）对少数类样本识别率较低。鉴于SVM+技术能利用样本间隐藏信息的启发，提出了多任务学习的不平衡SVM+算法（MTL-IC-SVM+）。MTL-IC-SVM+基于SVM+将不平衡数据的分类表示为一个多任务的学习问题，并从纠正分类面的偏移出发，分别赋予多数类和少数类样本不同的错分惩罚因子，且设置少数类样本到分类面的距离大于多数类样本到分类面的距离。UCI数据集上的实验结果表明，MTL-IC-SVM+在不平衡数据分类问题上具有较高的分类精度。

Multi-task learning of SVM+for imbalanced classification

When learning from imbalanced datasets, the traditional support vector machines(SVMs) had a low rate of identification on the minority class. Inspired by that SVM+ can utilize the additional information hidden in the training data and multi-task learning can improve the generalization performance by training multiple related tasks simultaneously, this paper proposed a new support vector machine called multi-task learning SVM+ for imbalanced classification(MTL-IC-SVM+). MTL-IC-SVM+ incorporated the multi-task learning framework into SVM+ to hand the problem of class imbalance by applying the different penalty factors to the data, especially, the margin between the hypersphere and the minority class was as large as possible. Experiments conducted on several UCI datasets show that the proposed methods lead to very encouraging results on imbalanced datasets.