Scaling Up ROC-Optimizing Support Vector Machines
Gimun Bae, Seung Jun Shin (Department of Statistics, Korea University, Seoul, Republic of Korea)
ROC-SVM最初由Rakotomamonjy提出,直接最大化了ROC曲线(AUC)下的面积,并在类不平衡的存在下成为传统二元分类的有吸引力的替代品。 然而,它的实际用途受到高计算成本的限制,因为培训涉及评估所有O(n^2)。 为了克服这一限制,我们开发了一种可扩展的ROC-SVM变体,该变体利用不完整的U统计,从而大大降低了计算复杂性。 我们通过低等级内核近似将框架进一步扩展到非线性分类,从而实现复制内核Hilbert空间的高效训练。 理论分析建立了一个错误绑定,证明了提议的近似值,合成和真实数据集的经验结果表明,拟议方法实现了与原始ROC-SVM的AUC性能,训练时间大大减少。
The ROC-SVM, originally proposed by Rakotomamonjy, directly maximizes the area under the ROC curve (AUC) and has become an attractive alternative of the conventional binary classification under the presence of class imbalance. However, its practical use is limited by high computational cost, as training involves evaluating all O(n^2). To overcome this limitation, we develop a scalable variant of the ROC-SVM that leverages incomplete U-statistics, thereby substantially reducing computational comp...